Abstract

Why is it that international ideological enemies—states governed by leaders engaged in deep disputes about preferred domestic institutions and values—are sometimes able to overcome their ideological differences and ally to counter shared threats, and sometimes they are not? Alliances among ideological enemies confronting a common foe are unlike coalitions among ideologically similar states facing comparable threats. Members of these alliances are perpetually torn by two sets of powerful contending forces. Shared material threats push these states together, while the effects of ideological differences pull them apart. To predict when ideological enemies are and are not likely to ally in the pursuit of common interests, it is necessary to know which of these contending forces is likely to dominate at a particular time. The values of two ideological variables beyond that of ideological enmity play the key role in determining outcomes: (1) states’ susceptibility to major domestic ideological changes and (2) the nature of the ideological differences among countries. Similar levels of ideological enmity and material threats will have vastly different effects on leaders’ alliance policies as the values of these additional ideological variables alter.

Introduction

The nature of alliances is one of the most important topics in the analysis of international relations. Yet, a major category of alliances has until now escaped scholars’ attention: cooperation between ideological enemies when these countries confront a common and pressing material threat to their security.1 This lack of systematic analysis is surprising for two reasons. First, the variation in the likelihood of alliance formation between ideological enemies—that is, states whose domestic institutions and values oppose each other—is high. On the one hand, there are many prominent examples of fierce ideological enemies allying to contain or defeat shared threats. Czarist Russia allied with republican France against Germany in the decades before World War I. The United States—a liberal, capitalist country—allied with the Soviet Union—a communist power—against Nazi Germany during World War II. Saudi Arabia, a theocratic monarchy, formed a tight alliance for most of the 1980s with Iraq, a secular authoritarian regime, against revolutionary Iran.

On the other hand, there also are numerous examples of ideological enemies not forming or preserving an alliance despite significant material pressures to do so. Britain and France, two liberal democracies, failed in the 1930s to ally with the Soviet Union despite the massive threat posed by Germany. The key leaders of the People's Republic of China before 1979 consistently refused to ally with the United States despite the clear danger posed by the Soviet Union.2 In 2009 and 2010, Turkey's Islamist government ruptured the country's alliance with Israel despite an increasing threat posed by Iran.3

A second reason why the lack of systematic analysis of alliances among ideological enemies is puzzling is because states’ leaders may find it difficult to predict when these coalitions are and are not likely to form and endure, sometimes underestimating or overestimating the probability of their doing so. In both cases, the results can prove disastrous for states’ interests. Prominent examples of both types of errors include German leaders’ belief in the decades before World War I that czarist Russia was unlikely to ally with republican France; British and French leaders’ opinion in the 1930s that Germany would not ally with the Soviet Union; U.S. leaders’ judgment in the early 1950s that Iran intended to ally with the Soviet Union despite Prime Minister Mohammad Mossadegh's opposition to communism; Soviet leaders’ belief beginning in the early 1970s that China had committed to an alliance with the United States; and the mistaken position among key officials in the George W. Bush administration that Iraq had allied with al-Qaida, a radical Islamist terrorist network.4

The unique nature of alliances among ideological enemies complicates the ability of analysts to judge the likelihood of these coalitions forming and lasting. Unlike coalitions among ideologically similar states facing comparable threats, alliances among ideological enemies are perpetually torn by two sets of contending forces. Shared material threats push these states together while their ideological differences pull them apart. Members of cross-ideological alliances are thus simultaneously friends and enemies, or “frenemies.”5

To predict when ideological enemies are likely to ally to confront a common threat, policymakers and scholars need to be able to identify which contending forces are likely to dominate at a particular time. No study, to my knowledge, has developed an analytical framework focused on this objective. This article seeks to fill the gap.

The core claim of my argument is that two ideological variables, in addition to the variable of ideological enmity, play the key roles in determining when either the rivalry or common-interest dimension of frenemy (or cross-ideological) relationships is likely to prevail. The first variable is the level of regime vulnerability in the members of a potential frenemy coalition. By “regime vulnerability,” I mean the susceptibility of a government to major ideological changes up to and including revolution.

The second variable is the particular configuration of ideological distances that exists among potential frenemy allies and their shared material threat. “Ideological distance” is the degree of ideological differences dividing the leaders of any two states.6 The more the ideological beliefs of the leaders overlap, the smaller the ideological distance; the less their ideologies overlap, the larger the ideological distance. “Configurations of ideological distances” refer to the various ways in which one state relates to at least two others along an ideological continuum. More specifically, configurations of ideological distances reflect whether the ideological distance separating a state and its potential frenemy ally is larger than, smaller than, or equal to the ideological distance with its material danger.

Levels of regime vulnerability and different configurations of ideological distances affect the two most important determinants of leaders’ willingness to pursue alliances with ideological enemies: the anticipated costs of committing to these alliances and the perceived need to do so. High levels of regime vulnerability significantly increase the costs to leaders’ core domestic interests created by cross-ideological alliances. When levels of regime vulnerability are high, leaders tend to worry that allying with an ideological enemy abroad will facilitate major ideological changes at home. This anticipated cost to leaders’ domestic interests will greatly reduce their willingness to join a cross-ideological coalition.

The configuration of ideological distances among potential frenemy allies and their shared material threat affects the perceived need for these states to ally. It shapes leaders’ understanding of the magnitude of international threats. For example, when a state is ideologically closer to a material threat than to a potential frenemy ally, it will feel less need to commit to a cross-ideological coalition. Ideological similarities with the material threat will tend to lower the perceived danger posed by this state, thereby offsetting the incentives for balancing against it created by material variables. In some cases, these ideological similarities may be so great as to convince leaders to ally with the material threat to counter the frenemy ally. In contrast, when a state is ideologically closer to a potential frenemy ally than it is to their shared material danger, its need to commit to the alliance will be enhanced. Large ideological differences with the material threat will push leaders to view this state as particularly aggressive and untrustworthy, thereby augmenting the incentives to balance against it.

This article proceeds in six sections. The first two sections describe the nature of frenemy alliances and why these coalitions are perpetually torn by powerful forces both for and against cooperation. In the third section, I identify and develop the conditions that determine when ideological enemies are most likely and least likely to pursue alliances against shared material threats. The fourth section presents the argument's predictions, and the fifth tests them in a case study that explores French leaders’ alliance policies toward the Soviet Union against Germany in the 1930s. I conclude with an examination of the argument's theoretical and policymaking implications for alliances among ideological enemies.

What Are Frenemy Alliances?

Frenemy alliances are coalitions of ideological enemies created by the existence of a shared, substantial material threat to these states’ security. I define “ideologies” as the principles of governance to which political leaders are dedicated. They reflect elites’ preferences for ordering the domestic-political world: the core institutional, economic, and social goals that they try to realize in their states.7 Key ideological issues include whether leaders support representative or authoritarian political institutions, capitalist or socialist economies, theocratic or secular values, and full rights of citizenship for some or all groups in their state. Prominent ideologies include communism, fascism, liberalism, monarchism, and religious fundamentalism.8

Ideologies such as those listed above can be divided into two broad types, and these types will have very different effects on leaders’ threat perceptions and international security policies, including alliances. Stephen Walt labels these types of ideologies “divisive” and “unifying.” Gregory Gause labels them “hierarchical” and “non-hierarchical.” Divisive or hierarchical ideologies call for political integration and the uniting of co-ideologues into a single country (e.g., pan-Islam) or “for members to form a centralized movement obeying a single authoritative leadership” (as was the case for international communism). Unifying or non-hierarchical ideologies, in contrast, do not prescribe political integration or single authoritarian leadership and, instead, recognize the legitimacy and sovereignty of the different states that share the ideology. Examples of unifying ideologies include liberalism, monarchism, and fascism.9

These different types of ideologies can produce two sources of ideological enmity. The first source exists when leaders are objectively devoted to opposing ways of ordering domestic politics in terms of core institutions and values. Relations among policymakers devoted to different unifying ideologies (e.g., liberals versus monarchists) and between one set of leaders dedicated to a unifying ideology and another set committed to a divisive one (e.g., liberals versus communists) fall into this grouping.

The second source of ideological enmity results when leaders in different states possess objectively similar ideological beliefs, but the ideology is a divisive one. When leaders subscribe to an ideology that does not respect state sovereignty or independence and instead calls for political integration or a single leader of a transnational ideological group, they will often come to view one another as intense ideological competitors who deny each other's legitimacy. Although shared devotion to a divisive ideology may result in some important ideology-based cooperation among leaders, the frictions created by this set of beliefs can make this cooperation difficult to sustain. As part of an effort to determine who should head the ideological group, leaders committed to divisive ideologies will often engage in fierce doctrinal debates over who possesses the best or purest interpretation of the ideology. Rivals in this debate will not view one another as ideologically similar, but as heretics or apostates. Such views existed, for example, among leaders of Catholic, Lutheran, and Calvinist states during the wars of religion in the sixteenth and seventeenth centuries and between the Soviet Union and China beginning in the 1960s.

Regardless of its source, ideological enmity can create powerful barriers to international security cooperation. International ideological relationships often play a major role in shaping how leaders judge others’ foreign policy intentions. Elites in different states who share a unifying ideology (i.e., high ideological similarities) tend to perceive one another as trustworthy individuals who share the same values and interests. Ideological enemies, in contrast, are likely to assume the worst about one another's international intentions. They view one another as untrustworthy and dedicated to opposing objectives. Most worrisome, leaders frequently believe that ideological enemies will use their states’ power in an aggressive manner, thereby endangering the security of their state.10 According to Charles Duelfer and Stephen Dyson, “The core of an enemy image is the assumption of malign intent. All behavior is seen as evidence of malign intent—with even cooperative-seeming behavior perceived as hostile—a function of either intent to deceive, a temporary weakness, or a retreat in the face of firmness from the perceiving state.”11 Ideological rivalry tends to result in the creation of these enemy images. The more fearful leaders are of one another's intentions, the more hostile their relations are likely to be.

History is replete with examples of policymakers explicitly attributing malign intentions to others based on ideological differences. President Dwight Eisenhower, for example, asserted in August 1954 that “the central core of the great world problem [the Cold War] is the aggressive intent of international communism.”12 Iranian leaders have consistently cited ideological enmity to explain their country's hostile relations with the United States. As Supreme Leader Ayatollah Ali Khamenei put it in 2003, “The primary reason for U.S. hostility toward our country is the Islamic identity of our system.”13

Ideological enemies that assume the worst about each other's international intentions are likely to greet the possibility of durable cooperation with tremendous skepticism. Former Soviet Foreign Minister Maxim Litvinov, for example, stated in 1946 that the “root cause” of the incipient U.S.-Soviet confrontation was “the ideological conception prevailing [in the Soviet Union] that conflict between the Communist and capitalist worlds is inevitable.”14

Under typical conditions, ideological enmity would be sufficient to prevent cross-ideological alliances. The tendency for ideological rivals to view one another as aggressive, untrustworthy, and possessing fundamentally opposed goals is not a firm foundation upon which to build extensive security cooperation. Frenemy relationships are unlike relations among most ideological enemies, however. While confronting ideological barriers to forming an alliance, frenemies also have substantial material incentives to ally. The next section discusses these material incentives for cooperation.

Why Ideological Enemies Ally

To explain why ideological enemies may form an alliance, I rely on insights from realist balancing theories, which provide the dominant explanation in the literature of alliances as means of aggregating capabilities against common dangers. To realists, states form alliances when they are sufficiently threatened by the combined effects of another state's material variables—specifically, aggregate power (based on economic and military capacity), geographical proximity, offensive capabilities, and displayed aggressiveness (these are the defining variables of balance of power and balance of threat theories). For realists, ideologies have minor effects on alliance policies, especially when the material threat is high.15

Based on realist understandings of threat and alliance formation, two ideological enemies will have strong incentives to form an alliance when confronting a geographically close state that is more powerful (especially in terms of offensive capabilities) and that state has indicated a willingness to use force to resolve conflicts of interest with the other two. Leaders are likely to prefer to pass the costs of balancing shared threats to others. Such buck-passing policies become increasingly risky, however, the greater the material danger. The greater the threat, the more powerful the incentives for leaders to counter it by balancing rather than buck-passing.

In all the examples I use to develop and test my argument, two or more ideological enemies confronted a shared, significant material threat that created strong incentives for forming an alliance. The existence of a shared, substantial material threat is what transforms an ordinary relationship among ideological enemies into a frenemy one, and it is a key scope condition of my argument.

This scope condition creates important analytic advantages for testing the argument. Holding material incentives for the formation and preservation of an alliance largely constant allows me to control for the effects of realist variables on alliance formation and stability. If there is marked variation in alliance outcomes despite the constancy of the initial scope condition of a clear material threat, then other factors must be responsible for this variation.16

At the same time, however, this scope condition has an important limitation. My argument applies best to situations when ideological enemies are in highly threatening environments created by shared material dangers, but they are not under attack. When states are under attack, realist balancing theories regarding the likelihood of ideological enemies allying are likely to be correct.

When Do Ideological Enemies Ally?

Levels of regime vulnerability and configurations of ideological distances that exist among the frenemy states and their shared material danger are instrumental in determining which set of forces—the material threat that pushes ideological enemies into an alliance or the effects of ideological enmity that pull them apart—will influence their alliance policies more. The variables of regime vulnerability and configurations of ideological distances are linked because they reflect how ideologies critically shape leaders’ assessments of threats to their two most important political interests: the security of their state and the stability of the regime type they support.

When regime vulnerability is high, leaders will tend to worry that an alliance with a country that shares the ideology of the domestic revolutionaries will increase the likelihood of regime change. This fear will greatly enhance the anticipated costs of the alliance.

Configurations of ideological distances refer to how a state relates to a potential frenemy ally and their shared material threat along an ideological continuum. More specifically, this variable indicates whether the ideological distance separating a state and its potential frenemy ally is greater than, less than, or equal to the ideological distance with the material threat. These variations in ideological distances shape how leaders understand the perceived danger posed by the material threat relative to the potential frenemy ally, and thus the extent of the need to form a cross-ideological coalition. Some configurations of ideological distances reduce the perceived danger posed by the material threat, thereby lowering the likelihood of frenemy alliances against it. Others amplify this danger, thereby making frenemy alliances particularly likely.

Because my argument concentrates on attributes that all ideologies and members of all ideological groups share, the predictions apply to all ideological relationships regardless of the ideologies under investigation. All leaders have a preeminent interest in maintaining their political power and the ideological order they champion. Therefore, variations in the level of regime vulnerability are likely to push members of all ideological groups to act in similar ways. Related analysis applies to the incentives created by specific configurations of ideological distances. As such, a system defined by relations among fascists, communists, and liberals in the 1930s will tend to exhibit dynamics similar to those of a system defined by relations among Catholics, Calvinists, and Lutherans in the sixteenth century. What matters is the type of ideological configurations that exist, not the specific ideologies that create them. By focusing on ideologies’ commonalities, I create a generalizable theory of ideologies that examines their systematic effects.

LEVELS OF REGIME VULNERABILITY

During periods of high regime vulnerability, elites believe that the domestic order they champion is susceptible to revolution.17 When leaders confront international ideological enemies and their level of regime vulnerability is high, the threats posed by those states that share the ideology of the domestic revolutionaries will be acute. Elites at these times are likely to believe that not only do international ideological enemies possess hostile intentions, but they will be powerful spurs to domestic ideological subversion, by which I mean the likely undermining at home of one set of ideological principles and the spread of a rival one.

There are three main sources of regime vulnerability: the existence of powerful revolutionary groups inside a country; the demonstrated power of revolutionary forces in other states as revealed by the overthrow or significant weakening of an existing government or governments; and the presence in the system of an ideological enemy that is trying to export its ideological principles. Powerful domestic revolutionary forces are the most important of these sources.18 These groups are a necessary condition for high levels of regime vulnerability because the other two sources of vulnerability intensify the effects created by these revolutionary parties. They are also a sufficient source of high levels of regime vulnerability because they will stimulate leaders’ fears of ideological subversion even if the other two sources are not in play.

When leaders confront strong material incentives to commit to a frenemy alliance but they are domestically vulnerable to the spread of the ally's ideology, their two most important political interests are in tension: allying with an international ideological enemy may enhance the security of their state while compromising the stability of their regime. For many leaders, the risk of domestic instability may be too great. High levels of regime vulnerability can augment the domestic costs of allying with ideological enemies by (1) increasing the likelihood that these alliances will empower and embolden domestic revolutionary groups; (2) enhancing the ally's ability to interfere in the vulnerable country's domestic politics; and (3) incentivizing domestic ideological mobilization campaigns directed against these alliances. On this last point, if an alliance with an ideological enemy is unpopular with elites’ core domestic supporters, then breaking this alliance will tend to please them.19 Because leaders will be most interested in mobilizing their base during periods of regime vulnerability, the incentives to weaken frenemy coalitions will increase at these times. When regime vulnerability is high, in other words, leaders are likely to adopt ideological policies—including ending alliances with ideological enemies—to rally their base against the domestic threat. The enhanced costs associated with frenemy alliances created by high levels of regime vulnerability in this situation are ones of opportunity. If leaders choose to commit to a frenemy alliance when breaking it would mobilize supporters against a pressing domestic threat, they would be forgoing significant domestic benefits.20

Powerful Domestic Revolutionary Forces. The greatest source of regime vulnerability results when leaders confront powerful revolutionary forces in their country, the most important of which I label “ideological fifth columns.” A “fifth column” is a domestic faction allied with a foreign country that undermines its government from within.21 Ideological fifth columns are defined by both domestic and international relationships. They are devoted to major ideological changes in their home countries while sharing the defining ideology of another state.

Ideological fifth columns can exist at both the societal and elite levels. At the societal level, they seek to promote an ideology shared by another state and are sufficiently organized and motivated to engage in mass protests involving thousands of people and/or an armed uprising against the government. As such, they are likely to increase leaders’ fears of regime vulnerability regardless of whether they govern a liberal or an authoritarian state. Because strong authoritarian governments possess a high capacity to suppress revolutionary movements, they may be able to better insulate themselves from revolutionary pressures than can either liberal regimes or weak governments.22 A high repressive capacity, however, will not necessarily alleviate fears of regime vulnerability.23 Even if a government has apparently crushed its revolutionary opponents, leaders’ levels of regime vulnerability are likely to remain high—likely for years—for several reasons. Governments can eliminate people but not ideas; powerful revolutionary movements have frequently reemerged after apparently being suppressed (e.g., liberals in monarchical France, communists in czarist Russia, and Islamists in the shah's Iran); and most importantly, when dealing with such critical interests as regime (and even personal) survival, elites confront powerful incentives to err on the side of caution and believe that even seemingly defeated revolutionary movements have a reasonable likelihood of reconstituting.

An elite ideological fifth column is a minority faction in the government. Elite fifth columns promote an ideology that is shared with another country and can influence policy, either negatively (by blocking the preferences of the dominant group or forcing it to adjust its policies in significant ways) or positively (by implementing its own agenda, at least in some key instances).24 Turkish Islamists in the Welfare Party in the mid-1990s were an example of this type of group. The Welfare Party controlled the office of prime minister in 1996 and 1997, sought to overthrow the strict secularist principles that defined the state, and was ideologically sympathetic to the ruling regime in Iran. Ideological fifth columns are not necessarily violent, but neither are they ordinary rival domestic parties. They are rival parties committed to vastly different ideological principles than a state's dominant decisionmakers and seek to replace the existing ideological order with their own.

Elite ideological fifth columns are particularly likely to stimulate leaders’ fears of revolution. The more political power an ideological rival possesses, the greater the likelihood that it will eventually dominate the government and transform the regime. More important for this article's purposes, ideological factionalization at the elite level is likely to provide foreign governments an enhanced opportunity to promote ideological change in the ideologically divided country because they have ideological allies in positions of power in this state. As Wade Jacoby finds, the ability of foreign states to influence others’ domestic politics is significantly amplified when target countries possess “substantial levels of internal contestation” in their governments, and when one of the contending political factions is “a minority tradition that favors the principles pushed by outsiders but that has, so far, failed to carry the day.” These conditions define elite ideological fifth columns. When they are met, “outsiders have a fighting chance of seeing their institutional preferences achieved” in the target state.25

High regime vulnerability created by the existence of ideological fifth columns increases the potential costs of alliances with ideological enemies by enhancing the risks associated with interactions with these states. The displayed power of these revolutionary forces makes it easier for foreign governments to identify potential sympathizers and provide them aid to achieve their objectives. The greater a foreign government's ability to subvert another's regime is, the more leaders of other states will want to minimize interactions with the foreign country out of fear that it will take advantage of its subversive potential.

These dynamics help explain the end of what became a frenemy alliance between China and the Soviet Union. Despite the obvious ideological similarities between the two communist powers, in the 1960s, Chairman Mao Zedong came to view the Soviet Union as a major danger to his domestic interests. Indeed, perhaps the most important cause of the Sino-Soviet split was Mao's belief that the Chinese Communist Party possessed a powerful faction dedicated to (in Mao's judgment) the illegitimate “revisionist” socialism created by Premier Nikita Khrushchev, rather than to Mao's more radical principles and objectives. Defeating this domestic challenge posed by the elite ideological fifth column, Mao believed, required increasing hostility toward the Soviet Union, which ended the countries’ alliance. As Mingjiang Li summarizes, “Mao could not show any leniency towards the Soviet leaders, because opposing international revisionism was part of his grand program for preventing revisionism in China. … [This calculation] completely destroyed the foundation of the Sino-Soviet Alliance.”26

Successful Revolutions Abroad by Ideological Rivals. Successful revolutions in other countries, which are a second source of regime vulnerability, will exacerbate leaders’ fears of domestic ideological change created by ideological fifth columns. According to the “demonstration effects” literature, when particular policies or actions succeed in one state, leaders or activists in others are likely to be inspired by these successes and try to emulate them.27

The transnational nature of ideologies suggests that ideological developments in one country will be replicated in others. Because individuals in different states can share ideological beliefs, ideologies can act as transmission belts that help connect developments across borders.

The success of any type of ideological enemy can generate demonstration effects that stimulate ideological changes in other states. Especially threatening to others’ domestic interests are ideological enemies that have recently come to power through a revolution. An ideological enemy that is also a revolutionary regime is particularly threatening to others because it represents the most abrupt and extreme threat to leaders’ core domestic interests and possibly their lives.

Scholars have found that the logic of demonstration effects applies to revolutions, which helps explain why revolutions and attempted revolutions often cluster in time, including those in Europe in 1848, the “color” revolutions in eastern Europe and central Asia in the 2000s, and the “Arab Spring” mass protests that began in 2010.28 Because revolutions spread, elites’ perceptions of regime vulnerability are likely to increase when these leaders witness their ideological enemies coming to power in other countries.29

Leaders’ fears of ideological contagion will likely last only as long as the rival ideology appears to be expanding. Analysts refer to the clustering of regime changes as “revolutionary waves”; historically, these have varied considerably in length.30 It is reasonable to assert that elites’ fears of ideological contagion will continue within a few years of the last revolution. After this time, leaders are likely to believe that the revolutionary wave has subsided and, with it, their fears of regime vulnerability.

Successful revolutions abroad can be connected to an increased likelihood of revolution at home in a number of ways. Most important, they will enhance the prestige and perceived viability of the revolutionaries’ ideology while damaging the prestige and legitimacy of the ideology that the revolutions replaced.31 These trends will help empower and embolden individuals throughout the system that share the ideology of the revolutionaries (i.e., members of ideological fifth columns). Enhanced ideological prestige will help empower a fifth column by encouraging new recruits. Growing ideological prestige will embolden a fifth column by pushing members to increase their estimated likelihood of revolutionary success, which will help spur them into action.

The boost that ideological fifth columns are likely to receive in revolutionary eras will increase the costs to leaders of committing to frenemy alliances. The more elites believe that the success of a competing set of beliefs is eroding the legitimacy of their ideology, the more they will want to isolate, stigmatize, and harm proponents of the rival ideology. Committing to an alliance with a state that shares the spreading rival ideology will impede these efforts. Indeed, committing to a frenemy coalition is likely to further enhance the competing ideology's prestige—and thus the likelihood of the ideology spreading—because alliance with an ideological enemy recognizes the legitimacy and value of this state. These beliefs and anticipated outcomes will reduce the likelihood of cross-ideological coalitions.

The preceding relationships help explain a reversal in Uzbekistan's alliance policies in the mid-2000s. Despite the major ideological gap separating Uzbekistan's authoritarian regime and the United States, Uzbek leaders enthusiastically welcomed U.S. offers of an alliance, which were designed to assist the United States in the war in Afghanistan. President Islam Karimov supported the alliance because it strengthened efforts to destroy terrorist Islamist groups (which also threatened Uzbekistan), aided in the balancing of Russian power, and provided annually hundreds of millions of dollars in aid.32

In 2005, however, a successful revolution in neighboring Kyrgyzstan (the “Tulip” Revolution), coupled with protests by Uzbek revolutionary forces (which revealed the existence of an ideological fifth column at the societal level) led to a major increase in Uzbek leaders’ perceptions of regime vulnerability. To President Karimov, revolution abroad was inspiring revolutionary forces at home. As Shahram Akbarzadeh explains, “There is little doubt that President Karimov saw [domestic protest] as an attempt to replicate the color revolutions.”33 Viewing close cooperation with the United States—a country that shared the ideology of the domestic protestors and Tulip revolutionaries—as increasingly threatening to his regime's core domestic interests, in July 2005 Karimov suspended the alliance, even though the factors that led Uzbek leaders to eagerly support the frenemy coalition remained.

Ideological Exportation by the Potential Frenemy Ally. Ideologies can spread from one country to another through more than demonstration effects. Countries also can exploit others’ domestic vulnerabilities by supporting ideological allies (fifth columns) in other states. Leaders can seek to export their ideology in ways that vary significantly in their forcefulness. At one end of the spectrum, leaders can try to boost the power of ideological allies in the target country by rhetorically delegitimizing the target's government, by giving financial support to the fifth columns, and by implementing economic sanctions against the target. At the other end of the spectrum, they can provide arms to revolutionary groups, assassinate officials, or launch an invasion.34

Attempts to export an ideology create obvious barriers and increased costs to cross-ideological coalitions, even in relation to ideological rivals engaged in soft forms of promotion. Leaders will be mistrustful of a state endeavoring to increase the power of ideological rivals in their country. The greater the mistrust of the ideological enemy is, the greater the perceived risks are of relying on this state to help protect their country's security. Elites in this situation are likely to highly doubt whether the frenemy ally will honor its commitments. An alliance with an ideological enemy that is endeavoring to export its ideology also is risky because close interactions with this state are likely to provide greater opportunities for it to engage in ideological proselytization and provide other forms of assistance to ideological fifth columns, thereby increasing these groups’ ability to affect revolutionary change.

Even if leaders who once engaged in ideological promotion renounce the policies, impediments to establishing cross-ideological coalitions are likely to remain high, because feelings of mistrust and resentment will not quickly dissipate. This mistrust will be particularly strong when the same individuals (as opposed to the same state) who once engaged in ideological promotion promise not to do so in the future.

Regardless of the cause of high levels of regime vulnerability, the effects on the probability of frenemy alliances will be the same. The more that leaders fear that allying with international ideological enemies will facilitate major ideological changes at home, the higher the anticipated costs of the alliance and thus the more reluctant such leaders will be to commit to these coalitions. For a summary of how the sources of regime vulnerability decrease the likelihood of frenemy alliances, see figure 1.

The Linkages between High Levels of Regime Vulnerability and Reduced Likelihood of Frenemies Allying

Figure 1.
The Linkages between High Levels of Regime Vulnerability and Reduced Likelihood of Frenemies Allying
Figure 1.
The Linkages between High Levels of Regime Vulnerability and Reduced Likelihood of Frenemies Allying

In contrast, when leaders’ perceptions of regime vulnerability are low, basing their alliance policies on the logic of realpolitik will be easier: the lower the level of regime vulnerability, the lower the potential domestic costs associated with committing to a frenemy alliance. When elites do not feel that they have to choose between the security of their state and the stability of their regime, they have more political space to act on the realist incentives pushing the ideological enemies together.

The likelihood of frenemies forming an alliance becomes even higher if these states’ leaders believe that their domestic interests are, to some degree, interconnected based on ideological commonalities, shared domestic threats resulting from mutual hostility to a third ideology, or both. For example, despite its profound ideological gap with the United States, China's domestic modernization campaign initiated at the end of the 1970s (led, most notably, by Deng Xiaoping) produced for the first time an important degree of ideological overlap. Before embarking on a tour of European states in May and June 1978, Deng told Vice Premier Gu Mu that “we ought to study the successful experiences of capitalist countries and bring them back to China.” The ideological agreement based on Deng's preference to incorporate “market elements into China's socialist system” incentivized cooperation based on China's need to learn from the economic successes of the United States.35 Combined with the material incentives for alliance that had existed throughout the decade, U.S.-China security interactions shifted from “the illusion of military relations” to an “alignment in confrontation against the Soviet Union.”36 As Deng reportedly told assistants after returning from a visit to the United States in early 1979: “If we look back, we find that all those [developing states] that were on the side of the United States have been successful [in modernizing], whereas all of these that were against the United States have not been successful. [Because we desire modernization], we shall be on the side of the United States.”37

Similar incentives for the creation of cross-ideological coalitions will exist when two ideological enemies are domestically vulnerable to a third ideology. In this situation, ideological enemies will have incentives to support the other's regime—despite their ideological differences—to inhibit the spread of a more pressing domestic threat. Mutual fear of revolution to political Islam, for example, was central to the tight alliance formed between Iraq and Saudi Arabia in the wake of the Iranian Revolution in 1979.38

CONFIGURATIONS OF IDEOLOGICAL DISTANCES

The second ideological variable critical to determining whether or not ideological enemies are likely to commit to an alliance to contain a shared material threat is the configuration of ideological distances among these states. In developing this variable, I stipulate a system of three powers: the initiating state, its potential frenemy ally, and their shared material threat. I then vary the ideological distances among these three countries from the initiating state's perspective based on whether (1) the initiating state views its potential frenemy ally or the shared material threat as a greater, lesser, or equal ideological enemy; (2) whether the initiating state and the material threat share an ideology; and (3) whether the potential frenemy ally and the material threat share an ideology. Based on these variations, five ideological configurations of how the initiating state relates to the potential frenemy ally and material threat along an ideological continuum are possible. These configurations significantly influence the likelihood of the initiating state agreeing to a frenemy alliance by affecting the perceived need to ally, most importantly by shaping the initiating state's assessment of how menacing the material threat is compared to the potential frenemy ally. Variations in configurations of ideological distances also affect the perceived need to form frenemy alliances by shaping the initiating state's assessment of how successful alterative security strategies to balancing and alliance formation—specifically buck-passing and band-wagoning—are likely to be.

The first three ideological configurations tend to decrease the perceived need for the initiating state to pursue a coalition with the potential frenemy ally, thereby offsetting realist reasons for cooperation. The fourth and fifth configurations create, paradoxically, ideology-based reasons for ideological enemies to ally, thereby augmenting the perceived need for alliance. In these configurations, cross-ideological alliances are even more likely than realist balancing theories predict. Variations in the perceived need to commit to a frenemy alliance across the five ideological configurations exist even though the realist incentives for alliance are identical in all. In all five, the existence of a shared material threat (based on an aggregated analysis of power distributions, offensive capabilities, geographical closeness, and displayed aggressiveness) creates strong incentives for the initiating state to commit to an alliance with the potential frenemy ally.39

To determine states’ relative positions along an ideological continuum, I place each pair of states in one of three broad categories: ideological enemies, ideologically similar states, or lesser ideological enemies in relation to greater ones. Ideological enemies condemn as illegitimate each other's domestic principles and institutions, describing them as ineffective, immoral, and in conflict with their own. Ideologically similar groups view each other as part of a transnational community based on a shared commitment to similar ways of organizing domestic politics. There can be noteworthy political differences within this community (liberals, for example, can support presidential and parliamentary democracies), but members will not denounce these differences as illegitimate. Lesser ideological enemies will condemn each other's principles and institutions. At the same time, they will be in ideological agreement on important issues, especially in relation to a third ideology. These lesser ideological enemies believe that a shared ideological rival represents a more profound challenge to their ideological beliefs than they do to each other. This shared ideological rival is thus the greater ideological enemy. Officials will tend to deem ideological enemy states as high threats based on assumptions of malign intent, ideologically similar states as low threats based on assumptions of benign intent, and lesser ideological enemies as in between these positions when compared to greater ideological enemies.

In all the examples I use below, leaders were clear in how they situated their state ideologically in relation to a potential frenemy ally and their shared material danger, thereby making coding of these configurations relatively uncomplicated. For a summary of how ideological configurations influence alliance preferences, see figure 2.

Linkages between Configurations of Ideological Distances and the Probability of Frenemies Allying

Figure 2.
Linkages between Configurations of Ideological Distances and the Probability of Frenemies Allying
Figure 2.
Linkages between Configurations of Ideological Distances and the Probability of Frenemies Allying

Ideological Betrayal (Inhibits Alliance). In the first ideological configuration, leaders in the initiating state are pushed for realist reasons to commit to an alliance with an ideological enemy (the potential frenemy ally) against a state (the material threat) that is ideologically similar to theirs (the initiating state and the material threat share a unifying ideology). I label this configuration “ideological betrayal.” An example of this dynamic occurred in the 1930s when fascist Italy was pushed to ally with liberal France (the potential frenemy ally) against a fellow fascist state, Germany (the material threat).

The dynamics of ideological betrayal are likely to lower the need for the initiating state to ally with the frenemy ally by significantly reducing the perceived danger posed by the material threat. The high ideological similarities uniting the initiating state and the material threat in this configuration will often result in high levels of trust and common interests that obviate the need for costly balancing policies, including allying with the potential frenemy ally. Indeed, major common interests created by a shared ideology will create strong incentives for the initiating state to ally (or bandwagon) with—not against—the material threat.

Italian leader Benito Mussolini clearly recognized the realist factors (which included Germany's massive power advantages and major conflicts of interest between Germany and Italy over Austria) that were pushing Italy to ally with France against Germany.40 Yet, the forces of attraction created by ideological similarities with Germany ultimately proved too powerful to overcome. Mussolini believed that high levels of ideological commonalities between Italy and Germany created extensive common interests, including the same international enemies (chiefly the Soviet Union and France, especially when led by the socialist-communist Popular Front coalition). As Robert Whealey summarizes, Germany's leader, Adolf Hitler, indicated to Mussolini that “he regarded ideology as the basis for an [alliance] agreement with Italy.” Mussolini responded in kind, including expressing in January 1936 his “great hope in creating a common front with Hitler against the democracy of England and France and against the communism of the Soviet Union.” The Italian dictator described the ideological foundations of Italian-German cooperation, which eventually resulted in an alliance, in a January 1936 meeting with a German emissary: “Between Germany and Italy there is a common fate. That is becoming stronger and stronger. … We have the same enemies, don't we? And Russia! This Russian Army—Bolshevism. Only we know about it. I and Herr Hitler.”41

Divided Threats (inhibits alliance). In the second ideological configuration, the initiating state is being pushed for realist reasons to ally with one ideological enemy (the potential frenemy ally) against another ideological enemy (the material threat), but the initiating state has more in common ideologically with the material threat. The material threat is a lesser ideological enemy, and the potential frenemy ally is a greater one. The central dynamic of this configuration is the divergence between the initiating state's greatest material and ideological dangers: the potential frenemy ally is the greatest ideological threat, and the material threat is the most pressing material danger. I label this configuration “divided threats.” For example, in the 1930s, British conservatives were being pushed to forge an alliance with the Soviet Union (the potential frenemy ally) against Nazi Germany (the material threat). Although both the Soviet Union and Germany were ideological enemies of Britain, British conservatives viewed the Soviet Union as the greatest ideological enemy in the international system. Fascist Germany was a lesser ideological enemy based on a shared commitment to anti-communism.

A configuration of divided threats will reduce the likelihood of cross-ideological coalitions. When states’ most pressing ideological and material threats diverge, there is likely to be confusion and disagreement about which country is the greatest overall menace. Is it the state (the material threat) with the greater material capabilities, but that shares some ideological beliefs with the initiating state? Or, is it the state (the potential frenemy ally) that is likely to be viewed as particularly untrustworthy and aggressive given its opposing ideology, but possesses less military power? The greater the uncertainty over the relative dangers posed by other countries, the lower the perceived need will be to commit to a coalition with the potential frenemy ally.42

The uncertainty and disagreement over threats that exists when states’ greatest material and ideological dangers diverge in a configuration of divided threats is illustrated by a debate in the British foreign ministry from 1937 to 1939, according to a summary by Zara Steiner, “over the relative dangers of the Fascist and Communist threats to the western democracies.” One faction argued that the Soviet Union—chiefly because of its particularly disparate ideological beliefs—was the greater threat. The other claimed that Germany was the chief danger primarily because of its greater extant capabilities and geographical proximity. Critically, “no consensus emerged [out of this debate], making it difficult to arrive at a settled policy,” especially whether to ally with the Soviet Union.43

Ideological Equidistance(inhibits alliance). In the third configuration, the initiating state is being pushed for realist reasons to ally with one ideological enemy (the potential frenemy ally) against another equally ideologically divergent one (the material threat), and these latter two states are themselves fierce ideological rivals. I label this configuration “ideological equidistance.” Here, the three states are ideological enemies at roughly the same level of intensity. China's relations in the 1970s with the United States and the Soviet Union, which Chinese leaders labeled “imperialist” and “socialist-imperialist” regimes, respectively, are an example of this configuration.

In a configuration of ideological equidistance, the initiating state is less likely to commit to a frenemy alliance because it believes that buck-passing will advance its interests. The hope for the state is that its ideological enemies will balance each other while it remains free of entangling commitments. During periods of ideological equidistance, policymakers in the initiating state will have good reason to expect that the material threat and potential frenemy ally will concentrate their hostilities on each other because of their own ideological enmity.

These dynamics help explain why Chinese leaders refused before 1979 to ally with the United States (the potential frenemy ally) against the Soviet Union (the material threat). The Chinese leadership should have been highly interested in allying their country with the United States for three reasons: in the 1970s, China was under very high threat from the Soviet Union; it was in more immediate danger from the Soviet Union than was the Untied States; and it was unprepared to counter this threat by itself. The effects of ideological equidistance, however, countered these incentives for alliance. Chinese officials, including Mao, were convinced that the ideological enmity between the superpowers would push them to focus their hostilities on each other. As an influential report submitted to the Central Committee by Chinese military leaders in July 1969 asserted: “The U.S. imperialists and the Soviet revisionists are two ‘brands’ of representatives of the international bourgeoisie class. On the one hand, they both take China as the enemy; on the other they take each other as the enemy.” Although the superpowers’ hostilities toward China were great, “the real threat is the one existing between themselves. … [Their ideological] contradictions [and thus] their hostilities toward each other are more fierce than ever before.”44

Chinese leaders’ confidence in the preeminence of the U.S.-Soviet rivalry created the hope that China's two most powerful rivals would bear the costs of containing each other. Internal documents reveal that Chinese leaders’ primary objective of improving relations with the United States in the early 1970s was not to set the stage for the creation of a balancing coalition, but to take advantage of and amplify U.S.-Soviet enmity so that the two superpowers would balance each other while China remained free from costly commitments. An April 1973 confidential report issued by the political department of the People's Liberation Army in Yunnan Province, for example, recommended “using one enemy against another (without, however, allying with either), to ‘aggravate the contradictions between the United States and the Soviet Union.'”45

In contrast, the fourth and fifth configurations of ideological distances among potential frenemy allies and their shared material threat facilitate the formation of an alliance. Here international ideological configurations, paradoxically, incentivize cooperation among ideological enemies. These configurations add to material incentives pushing ideological enemies to ally, thereby making frenemy coalitions even more likely than realist balancing theories predict.

Double threat (facilitates alliance). In the fourth configuration, the initiating state is being pushed for realist reasons to ally with an ideological enemy (the potential frenemy ally) against a state (the material threat) that is a greater ideological rival than the potential ally. In this configuration, the material threat is a “double threat” because it is the greater material and ideological danger to the initiating state. The initiating state has no, or only minimal, ideological commonalities with the material threat. In contrast, the potential frenemy ally and initiating state possess an important area of ideological agreement, making the former less of an ideological rival. During the Cold War, for example, Saudi Arabia was pushed for realist reasons to ally with the United States (the potential frenemy ally) against the Soviet Union (the material threat). Although the ideological gap separating the Saudi regime from the United States was considerable, Saudi leaders viewed atheistic communism as a greater ideological danger than liberalism.

In a configuration of double threat, ideological and material variables work together to make the material threat especially dangerous to the initiating state. The state with the greatest capacity to harm will also likely be viewed by the initiating state as particularly hostile and untrustworthy because of its disparate ideological beliefs. The greater the perceived threat is, the stronger the need for extensive balancing policies will be, including allying with (lesser) ideological enemies.

The incentives created by a configuration of double threat help explain the durability of the U.S.-Saudi frenemy alliance during the Cold War. Rather than preventing a U.S.-Saudi coalition, ideological calculations based on Saudi and U.S. leaders’ mutual hostility to communism powerfully reinforced the realist reasons pushing for it. According to Rachel Bronson, “Because Soviet-inspired Communism was based on a hostility toward religious belief, the more religious a country, the more likely it would be to rail against Communism and look toward the United States. … Saudi Arabia, a deeply religious state, was [for the United States] the perfect prophylactic against the spread of Communism and [thus] a natural American partner.”46

Ideological outsider (facilitates alliance). In the fifth configuration, the initiating state is being pushed for realist reasons to ally with one ideological enemy (the potential frenemy ally) against another (the material threat), both of which have similar ideological beliefs. In this configuration, the initiating state is an “ideological outsider,” ideologically isolated from the material threat and potential frenemy ally. France's relations with Germany and Russia in the decades before World War I are an example of this ideological configuration: realist calculations were strongly pushing republican France to ally with Russia (the frenemy ally) against Germany (the material threat), but France was an ideological outsider, and Russia and Germany were ideologically similar (both were monarchies).

When the initiating state is an ideological outsider, it and its potential frenemy ally confront a shared material threat, but the initiating state's leaders are likely to doubt that the potential frenemy ally will reliably help counter it. The initiating state's leaders are instead likely to anticipate cooperation between the other two countries because they are ideologically similar. Buckpassing in this situation would be risky. To attempt to free ride when others are more likely to cooperate with, rather than balance against, one another is likely to reinforce the initiating state's isolation and thus the dangers it confronts. The initiating state's concerns will increase its need to ally with the potential frenemy ally, both to balance the material threat and to ensure that the material threat and potential frenemy ally remain divided.47

The initiating state's objective of allying with the potential frenemy ally will be difficult to realize, however. When the initiating state is an ideological outsider, it has increased incentives to ally with the potential frenemy ally. The potential frenemy ally, though, will be reluctant to commit to this coalition because it will typically be operating in a configuration of ideological betrayal given its ideological similarities with the material threat. Thus, for the potential frenemy ally to ally with the initiating state against the material threat, it will have to set aside both repellant forces created by ideological disputes with the initiating state and attractive forces created by ideological similarities with the material threat. Given these twin barriers to cross-ideological alliance from the potential frenemy ally's perspective, the initiating state needs to be particularly aggressive in reaching out to the potential frenemy ally, likely even offering extensive security commitments. Being an ideological outsider, paradoxically, creates ideology-based incentives for the aggressive pursuit of alliances with ideological enemies.

The origins of the Franco-Russian alliance in 1894 from France's perspective illustrate these dynamics. French elites were convinced that the ideological similarities uniting Russia (the potential frenemy ally) and Germany (the material threat) would be a powerful source of cooperation between the two monarchies.48 To create a frenemy alliance with Russia while preventing a German-Russian rapprochement based on these states’ ideological similarities, French leaders believed that they had to be especially aggressive in their outreach, committing not just to the defense of Russia, but to the support of Russian offensive aims in the Balkans. As Georges Michon summarizes, to “detach Russia from Germany,” the French “had to offer sufficient inducements to the Tsar to compensate him for the loss of German friendship. These concessions … completely reassured the Tsar as to his interests in the Near East being safeguarded since France … would afford armed support for Russia's Balkan ambitions.”49

Predictions about the Likelihood of Frenemy Alliances

This section predicts the likelihood of cross-ideological coalitions when the effects of the two independent variables are combined. As explained, the two main pathways that connect the independent variables to changes in the probability of leaders pursuing frenemy alliances are the anticipated costs and the perceived need to create such coalitions. High levels of regime vulnerability in relation to the potential frenemy ally's ideology make the government's anticipated domestic costs of committing to this alliance high. Low levels of regime vulnerability result in low domestic costs of committing to the alliance. When a state's configuration of ideological distances with the potential frenemy ally and material threat are ideological betrayal, divided threats, or ideological equidistance, the state's ideology-based need to commit to a frenemy coalition is low. This perception offsets to a significant degree the incentives for alliance created by shared material threats, which are pushing for the creation of frenemy coalitions in all instances. When a state's ideological configuration with the potential frenemy ally and material threat is either one of double threat or ideological outsider, its ideology-based need to commit to the frenemy alliance is high, which augments the material incentives working for this outcome. The predictions based on different combinations of these variables are summarized in figure 3.

Predictions about the Likelihood of Frenemy Alliances

Figure 3.
Predictions about the Likelihood of Frenemy Alliances
Figure 3.
Predictions about the Likelihood of Frenemy Alliances

The predictions in cells 1 and 4 in figure 3 are straightforward and compelling. In cell 1, low levels of regime vulnerability and ideological configurations of double threat or ideological outsider make the domestic costs associated with joining a frenemy alliance low and the ideology-based need to join high. The incentives created by these outcomes make the pursuit of a frenemy alliance the dominant outcome. When costs are low and the need is high, leaders are likely to seek to ally their state with an ideological enemy to counter a shared material danger. France's alliance with czarist Russia before World War I exemplifies these dynamics.50

In cell 4, high levels of regime vulnerability and ideological configurations of ideological betrayal, divided threats, or ideological equidistance make the domestic costs associated with joining a frenemy alliance high and the ideology-based need to join low. In this situation, a frenemy alliance is unlikely. When the costs to vital interests created by allying are substantial and the perceived need to adopt this policy in significant doubt, leaders are unlikely to opt for this choice. French conservatives’ and Radicals’ intense aversion to allying with the Soviet Union for most of the 1930s illustrates these dynamics, as I discuss in the case study.

Predictions in cells 2 and 3 are more complicated, because the incentives created by the argument's two independent variables are pushing in different directions. In cell 2, ideological configurations of ideological betrayal, divided threats, or ideological equidistance make the initiating state's ideology-based need to commit to a frenemy coalition low, while low levels of regime vulnerability make the likely domestic costs created by close ties with an ideological enemy small. Core dynamics of this cell for the initiating state are high uncertainty over threats and high flexibility in alliance partners. Uncertainty over threats is high because material variables push for active balancing against the material threat while ideological configurations offset this need. Alliance flexibility also is high because the domestic costs created by allying with ideological enemies are low. When this is the case, leaders’ alliance choices are less constrained by the effects of ideological differences.

Here, “hedging” policies are likely to be the dominant outcome. Cheng-Chwee Kuik defines hedging as “insurance-seeking behavior under high-stakes and uncertain situations, where a sovereign actor pursues a bundle of opposite and deliberately ambiguous policies vis-à-vis competing powers to prepare a fallback position should circumstances change.”51 Hedging policies are defined by their ambiguity and flexibility of partners, which match the alliance incentives of cell 2.

Applied to my argument, a hedging state would not form a tight alliance with either the potential frenemy ally or the material threat. It would instead opt for more ambiguous policies, including making supportive comments regarding the interests of both countries and engaging in security cooperation with both. The goal of these hedging policies is to buy time to determine if either the material threat or potential frenemy ally becomes a more certain danger that requires more forceful and definitive balancing. German Chancellor Otto von Bismarck's offers in the 1880s of security cooperation to both Russia (an ideologically similar regime but a material threat) and Britain (a potential frenemy ally) to restrain both states from threatening Germany's interests are an example of hedging policies in the situation I describe.52

The dynamics of cell 3 are the most complex and therefore the most difficult to predict. In this cell, the trade-offs between leaders’ two most important political interests—the protection of the security of their state and the preservation of their political power and the regime type they support—are particularly stark. High levels of regime vulnerability are likely to make the anticipated domestic costs associated with frenemy alliances substantial. The effects created by configurations of double threat and ideological outsider, however, will make the ideology-based need to commit to these coalitions strong. Given these competing forces, leaders operating in cell 3 are likely to be convinced of the need to commit to a frenemy alliance to protect their state's security, but will be fearful that this alliance will subvert their regime. Because of the extreme importance that leaders place on protecting both the security of their state and the continuation of their regime, it is unclear which of these outcomes they will tend to privilege. Alliance policies will depend on which set of threats (to the state or to the regime) leaders deem greater and more pressing.

France's Failure to Ally with the Soviet Union in the 1930s

In this section, I explain why liberal France did not form a stable frenemy alliance with the Soviet Union against Nazi Germany in the 1930s. I chose this case for two main reasons. First, it satisfies my scope conditions. The case involves relations among ideological enemies being strongly pushed by material factors to ally against a common danger in a system not yet in war. The key indices of threat, according to realist balancing theories, show that the danger posed by Germany to France was massive, making for powerful forces pushing for a Franco-Soviet alliance. Germany had a substantial power advantage over France; it is geographically close (including sharing a land border); and the two countries had important conflicts of interest for which German leaders had indicated a willingness to use force to resolve. By 1937, as Peter Jackson explains, French leaders were convinced that “Germany had established decisive superiority over France both on the ground and in the air. … The result was an often crushing sense of inferiority.”53

Second, the case involves substantial variation in the argument's independent variables. For France's conservative and centrist parties, fears of communist revolution in France (high regime vulnerability) and the belief that the Soviet Union was a greater ideological threat than Nazi Germany (a configuration of divided threats) worked against committing to a frenemy alliance with the Soviet Union. For conservatives and centrists, the costs of this alliance were high and the need for it was in significant doubt. For French socialists, in contrast, relatively low fears of communist revolution (low regime vulnerability) and the belief that Nazi Germany was the greatest material and ideological danger (a configuration of double threat) resulted in strong support for an alliance with the Soviet Union. For socialists, the need for the alliance was high while the costs were low. Adding to this variation, the value of one of the independent variables—regime vulnerability—changed significantly for conservatives and centrists halfway through the decade. These variations allow for a test of the argument's predictions when a host of other variables, especially those associated with realist theories, are held largely if not completely constant.

CONSERVATIVES’ AND RADICALS’ OPPOSITION TO A FRANCO-SOVIET ALLIANCE

The most powerful policymakers in France from 1934 to 1940 belonged to conservative and Radical-Socialist, or Radical, parties (the Radicals were the dominant centrist party).54 These groups’ alliance policies toward the Soviet Union can be divided into two distinct phases. In 1934 and 1935, a majority of France's most powerful Radical and conservative politicians and military leaders supported forming a frenemy alliance with the Soviet Union, consistently supporting their position with realist logic.55 Radical and conservative leaders in these years repeatedly referred to the grave material danger created by Germany's rise and how an alliance with the Soviet Union—because of its power and geography—was necessary to counter this threat. As Minister of State Edouard Herriot (a Radical) explained in 1935: “I consult the map. I see only one country which can bring us the necessary counterweight [against Germany] and create a second front in case of war. That is the Soviet Union.”56 In 1934 and 1935, Radical and conservative politicians took the lead in negotiating the Franco-Soviet Pact of Mutual Assistance. According to the Pact, if either country were attacked, the other would “immediately” provide it “aid and assistance.”57 The Pact was signed in May 1935 and ratified by the French government in February 1936.

If the signing and ratification of the Pact of Mutual Assistance were the final decisions on allying with the Soviet Union before the outbreak of World War II, it would be reasonable to conclude that France and the Soviet Union had forged an alliance and that realist calculations had superceded the effects of ideological enmity.

The end of 1935, however, marked a transition point in French politics and the beginning of the second (and highly ideological) phase in conservatives’ and Radicals’ alliance policies toward the Soviet Union, a phase that lasted until the end of the decade. Despite their high interest in allying with the Soviet Union in 1934 and 1935, both conservative and Radical elites reversed course after 1935 and became adamantly opposed to the coalition. By the time the Pact of Mutual Assistance was up for ratification, conservative politicians had overwhelmingly rejected it, with most refusing to vote for the accord that they had championed only shortly before.58

Almost all Radical members of the French parliament voted to ratify the Pact. Their support, along with unanimous consent from socialist and communist members, resulted in passage. By the fall of 1936, however, most Radicals—including then Minister of National Defense and War Edouard Daladier and Foreign Minister Yvon Delbos—had joined with conservatives to rebuff repeated Soviet requests to develop the Pact in ways that would have significantly increased its utility. Radical politicians and conservative leaders in the military refused to engage in concrete military cooperation with the Soviets, including staff talks, joint training, and war planning.59 Conservative and Radical elites also repeatedly questioned the Pact's value to the point where many—including Soviet leaders—doubted that France was committed to the agreement. After meeting with high-ranking Soviet officials in October 1938, France's ambassador to the Soviet Union, Robert Coulondre, told his superiors that, from the Soviet perspective, the Pact was “null and void.”60

Even in 1939, when the probability of war with Germany as judged by French intelligence was becoming very high, key French elites remained reluctant to commit their country to a military alliance with the Soviet Union that would have required extensive coordination with Soviet armed forces.61 Despite the urgency of the situation, French representatives sent to the Soviet Union in August 1939, who along with their British counterparts were ostensibly trying to finalize an alliance with the Soviet Union, lacked the authority to sign a military accord. When the Soviets asked “about the strength of the British and French forces, what plans had been made to fight the Germans, and how the western forces would be deployed,” the representatives could not respond with specifics, because “neither the British nor the French were prepared to discuss [with the Soviets] military cooperation in wartime.”62

HIGH VULNERABILITY AND HIGH COSTS OF ALLYING WITH THE SOVIET UNION

Changes in levels of regime vulnerability for conservatives and Radicals resulting from increases in the power of an ideological fifth column tied to the Soviet Union, the French Communist Party, were the factor most responsible for these leaders’ opposing alliance policies before and after 1936. Before 1936, the power of the communist fifth column at both the elite and societal levels was weak, which made its ability to shape governmental policy negligible. The Communist Party won only ten seats in the 1932 parliamentary elections. Popular protests in support of French communists’ political objectives were also relatively rare in this period, as were strikes on behalf of working-class interests.63

Critically, the weakness of the French Communist Party facilitated conservative and Radical support for an alliance with the Soviet Union. Members of the French right, as William Scott explains, “would accept the Franco-Soviet Pact so long as the French ‘Soviets’ [French communists] were only a nuisance” and not a powerful revolutionary threat.64 As long as regime vulnerability was low because of the weakness of the ideological fifth column, French conservatives and Radicals did not have to choose between the security of their state and the stability of their regime, which greatly facilitated their ability to base their alliance policies on the logic of realpolitik.

The power of France's communist fifth column at both the elite and societal levels increased considerably in the mid-1930s and, with it, conservatives’ and Radicals’ sense of regime vulnerability and consequent opposition to a Soviet alliance. French communists beginning in 1935 became members of a political coalition, known as the Popular Front, that promised to significantly increase their party's electoral fortunes. The Communist and Socialist Parties had agreed to the coalition in 1934 and the Radical Party, chiefly spurred by a fear of the rise of French fascism, agreed to join in 1935. (Radicals at this time considered French communists the lesser of two evils compared to French fascists. This view, though, would not last long. By the summer of 1936, most Radicals viewed communists as the greatest domestic revolutionary threat.)65

Communists’ participation in the Popular Front promised a significant boost to the party's power. The coalition was likely to win the 1936 parliamentary elections, and would greatly enhance the party's ability to shape policy, thereby making the party a powerful elite ideological fifth column.

Communists’ participation in the Popular Front was sufficient to stoke conservatives’ fears of regime vulnerability and, with them, opposition to forming an alliance with the Soviet Union.66 A Soviet alliance, conservatives argued, would provide French communists more prestige and the Soviet Union enhanced opportunities for ideological proselytization. Such outcomes were too risky to allow during a period of high regime vulnerability. As one conservative parliamentary member, Henri de Kerillis (who had advocated for a Franco-Soviet alliance before 1936), stated during a February 1936 debate in the Chamber of Deputies over ratification of the Pact of Mutual Assistance: “The Communists, in as much as they are active elements of the Popular Front, are trying to gain power, and we cannot combat them while at the same time proclaiming our friendship for M. [Soviet leader Joseph] Stalin. … At the Right, nationalist deputies have been declaring: ‘What a tragedy to have to vote against a possible ally [the Soviet Union] even if that ally is monstrous, at the moment when the German menace is becoming real!’ But we do not hesitate to say here that our friends have done well to vote in a body against the pact, considering the conditions in which it was presented and the [domestic] risks that it involves at present.”67

Although the mere likelihood of a communist ideological fifth column participating in a governing coalition was sufficient to give rise to conservatives’ fears of regime vulnerability to the point where they opposed allying with the Soviet Union, the aftermath of the June 1936 parliamentary elections powerfully reinforced these fears. The results of the election led to socialist leader Léon Blum becoming prime minister from June 1936 to June 1937. The assumption of power by the Popular Front was accompanied by a massive increase in the number of sit-in strikes across France, as more than a million factory workers occupied their places of work and demanded wage increases and social reforms.68 The Communist Party was responsible for organizing many of these strikes and protests.69

The developments described above, which indicated the existence of a powerful ideological fifth column at both the elite and societal levels, significantly increased fears of revolution among not only conservatives, but also Radicals.70 These fears pushed Radical leaders to repudiate their previous policies and oppose a cross-ideological coalition with the Soviet Union. Radicals believed that to ally with the Soviet Union at a time when the Communist Party was both increasing its power and flexing its political muscles risked further empowering of domestic revolutionary forces. Ambassador Coulondre, on behalf of Foreign Minister Delbos (a Radical), warned Soviet Foreign Minister Maxim Litvinov in November 1936 that Franco-Soviet relations were in a “critical” state because of the “position of the French Communist party.”71 In the foreign ministry's view, according to Nicole Jordan, “extensive reliance on the Soviets could only lead to uncontrollable ideological contagion [in France] and conflict with Germany.”72

The more that French leaders feared that cooperation with the Soviet Union would facilitate a communist revolution in their state, the more costly an alliance with the Soviets became. This relationship significantly reduced conservatives’ and Radicals’ willingness to commit to the coalition despite the ever-increasing German material threat. Prime Minister Blum summarized these relationships in a November 1936 letter to Ambassador Coulondre: “A psychosis is being created according to which the Soviet entente leads to Communism; this fear tends to neutralize that which is inspired by the German threat and to paralyze cooperation among the pacific powers at the very time when this current ought to intensify.”73

DIVIDED THREATS AND REDUCED NEED TO ALLY WITH THE SOVIET UNION

An ideological configuration of divided threats for conservative and Radical leaders further reduced the likelihood of a Franco-Soviet frenemy alliance in the second half of the 1930s. These leaders recognized that Germany was France's greatest material threat. Yet, given their belief that communism was the most radical departure from their preferred domestic ideological order, they viewed the Soviet Union to be the greatest ideological danger. Despite major ideological differences with Nazi Germany, conservatives and Radicals shared an important area of ideological agreement with this state: intense anti-communism.

The divergence between the greatest material threat and greatest ideological dangers that is the core of a configuration of divided threats reduced Radicals’ and conservatives’ willingness to commit to a Soviet alliance by helping to mask the extent of the German threat, thereby lowering the perceived need to ally with the Soviet Union. Mutual dedication to anti-communism, in other words, made Germany a smaller danger than an analysis based on material variables indicated—hence, the frequent description of Germany as a useful “bulwark” against communism.74 In the words of Chief of the General Staff Maurice Gamelin, anti-communism “made many of us lose sight of the dangers of Hitlerism and fascism at our doorstep because behind the ‘Popular Front’ one saw the specter of Bolshevism. Therein lies the origins of the slogans that disfigured the soul of the nation: ‘Better Hitler than Stalin’ and ‘Why die for Danzig?'”75 The lower the perceived threat posed by Germany given a shared commitment to anti-communism, the weaker the incentives pushing France to ally with the Soviet Union.

FRENCH SOCIALISTS’ SUPPORT FOR A FRANCO-SOVIET ALLIANCE

Offering additional support for my predictions, French socialists—led by Blum—pushed for an alliance with the Soviet Union throughout the second half of the 1930s, including during the year Blum was prime minister. According to Joel Colton, “Russia played a key role in the ‘grand design’ which [Blum] envisaged in order to counterbalance the weight of Germany and Italy. … The end result he contemplated was a triple partnership [among Britain, France, and the Soviet Union]—'a combination reproducing the Triple Entente of the years before 1914.'”76 Blum and fellow socialists not only consistently insisted that the Pact of Mutual Assistance be honored, but endeavored to strengthen it. Blum did so by ordering the start of General Staff negotiations with the Soviets and by working to create a triple alliance among Czechoslovakia, France, and the Soviet Union.77

For French socialists, the values of the two ideological variables that affect the probability of leaders pursuing frenemy alliances were different than those of conservatives and Radicals. These variables, as a result, added to the material incentives pushing for an alliance with the Soviet Union. Particularly important to socialists was the double threat configuration. To socialists, a fascist Germany was both the greatest ideological and material threat.78 This intense ideological enmity pushed French socialists to believe the worst about Germany's intentions. “Blum's international policy,” as Nathanael Greene summarizes, “was rooted in the belief that the inherently aggressive character of the German and Italian regimes constituted an imminent danger to the peace of Europe.”79 Given Germany's double threat, allying with the Soviet Union was the obvious policy choice.80

ALTERNATIVE EXPLANATIONS FOR ALLIANCE FAILURE

Other potential explanations for France's failure to ally with the Soviet Union exist, but none is as convincing as my argument. The most prominent are realist, buck-passing arguments that examine the effects of Europe's multipolar structure combined with widespread perceptions of defense dominance.81 In these accounts, power multipolarity reduced the incentives for France to ally with the Soviet Union by providing the hope that the Soviet Union would bear the costs of containing Germany while France remained a nonbelligerent. Perceptions of defense dominance were a key factor in pushing French leaders to prefer buck-passing to making alliances. The more secure a state is based solely on its own defenses, the lower the need to seek international allies.

There are two major problems with this explanation of alliance failure. First, French leaders were not sufficiently confident in the power of defensive weapons and France's ability to free ride on the balancing efforts of others to the point where they believed allies were unnecessary to the country's security. To the contrary, French elites throughout the 1930s were intensely interested in forming alliances, believing that France's defenses alone were inadequate to counter German aggression. Nor were they sufficiently confident that other states would be able to contain Germany while France remained free of commitments. If Germany did pursue aggressive action, geography dictated that it would attack in the West first, just as it did in World War I. So critical were alliances to French leaders’ war planning despite perceptions of defense dominance that, according to Barry Posen, they in some ways placed greater emphasis on external rather than internal balancing: “French politicians concentrated their balancing efforts abroad, rather than at home … because without powerful allies France could not expect to prevail in an attrition war with Germany. … The defense might have the advantage, but in a one-to-one slugfest Germany's large population and resource superiority might overwhelm French defenses”—hence, conservatives’ and Radicals’ strong interest in allying with the Soviet Union before 1936 and socialists’ push for this coalition for most of the decade.82

The second major puzzle for realist buck-passing explanations of France's failure to ally with the Soviet Union is the variation after 1936 of French leaders’ alliance preferences along party and ideological lines. This variation means that the key to understanding why France failed to form an alliance with the Soviet Union against Germany in the second half of the 1930s is not discerning why French leaders, in general, opposed this coalition, but why most conservatives and Radicals were hostile to the alliance whereas most socialists strongly supported the alliance. Realist buck-passing arguments cannot explain this variation. Because these arguments’ causal variables—including power polarity, power distributions, geography, the offense-defense balance, and other countries’ policies—are identical for all members of the same state, they cannot explain systematic variation in security policies by different ideological groups within a country.

The only major exception to France's partisan variation in alliance policies toward the Soviet Union occurred in 1934 and 1935, but even this outlier is puzzling for realist buck-passing arguments. French conservatives and Radicals in these years endeavored for their state to ally with the Soviet Union, not pass the buck to it. They opted for this choice even though France's power in these years in relation to Germany was at its most advantageous point in the decade. If realist buck-passing arguments offer the best explanation of France's policies toward the Soviet Union, one would expect French policymakers to have been the most prone to free ride on the balancing efforts of others when France's relative power position in relation to Germany was at its highest level. Conservatives and Radicals then switched from balancing to buck-passing when French relative power experienced deep decline. This, again, is contrary to realist predictions. States should be less, not more, likely to pass the buck as material threats to their interests increase.

Conclusion

When predicting the likelihood of cross-ideological alliances, it is tempting to believe that the famous realist dictum “the enemy of my enemy is my friend” provides an adequate guide. There is certainly much wisdom in this phrase. Mutual animosity toward pressing dangers is often sufficient to induce alliances among states, even among those founded on opposing ideological principles. The necessities of politics, including international politics, frequently result in strange bedfellows.

Unfortunately for ease of analysis and prediction, frenemy alliances defy easy categorization. While ideological enemies sometimes set aside their ideological disputes in the face of shared threats, leaders in these situations just as frequently adhere to the reverse logic, with the enemy of their enemy remaining an enemy.

The values of two ideological variables in addition to ideological enmity —the level of regime vulnerability for a country's key leaders and the nature of the configurations of ideological distances among potential frenemy allies and their shared material threat—provide the key to understanding when either the common-interest or rivalry dimension of frenemy relationships is likely to determine leaders’ alliance policies.

The article makes three main sets of contributions to the development of international relations theory. First, it adds to the literature on alliances by explaining an important category of coalitions that has heretofore not received systematic attention. This lack of attention is surprising given that leaders frequently misjudge the likelihood of frenemy alliances, often with very costly consequences, and that the number of possible frenemy coalitions is substantial given the high levels of ideological heterogeneity of many of the world's regions. The greater the number of ideological divisions among states, the greater the number of possible frenemy alliances, and thus the more likely it is that the nature and effectiveness of international balancing will depend on whether these types of coalitions are likely to form.

Second, the argument contributes to important literatures that explore the international security effects of domestic variables. The article adds to the literature on ideologies and international relations by demonstrating the centrality of ideological variables to states’ core security policies even in the least likely of circumstances: periods of high material dangers. It also contributes to the literature that demonstrates the importance of leaders’ domestic security to their international balancing policies.83 My analysis of regime vulnerability indicates that leaders’ alliance policies will frequently hinge on the state of domestic affairs and the likelihood that international ideological threats will augment domestic ones. The analysis thus points to an important security dilemma that realist analyses miss: in conditions of high regime vulnerability, a key means by which a country can increase its external security (allying with an international ideological enemy) is likely to undermine its internal security. Failure to consider the implications of this dilemma will result in an exaggeration of the likelihood of effective international balancing.

Third, the argument creates a framework for integrating the insights of realist and ideological theories on alliances. The analysis combines the effects on alliances created by material and ideological variables and then identifies the conditions in which either set of variables is likely to determine outcomes. The argument, as a result, can explain why the same states, and even the same leaders, can pursue highly ideological foreign policies in some instances and ones that conform with realist prescriptions in others.84

In some cases—such as when a country is under attack—an analysis based on the effects of material variables alone will typically be sufficient to determine which state is another's primary enemy and which policies (active balancing, including allying with ideological enemies) can best counter this danger. In many cases, though, ideological variables combine with material ones to determine primary enemies and appropriate responses to them. I have concentrated on one set of responses to international threats: when leaders are and are not likely to ally with ideological enemies against common dangers. The implications of the analysis, though, are even broader. The argument indicates not only when leaders are likely to ally to balance a material threat, but when they are likely to bandwagon with this state and when they are likely to try to pass the costs of balancing to other countries. At its broadest level then, the argument seeks to explain when states engage in a range of security policies when taking ideological variables into account.

The argument can also aid policymakers by helping them predict the likelihood of effective balancing both regionally and globally.85 Because, for example, leaders of Asian countries that would be of high utility in countering a rising China have indicated either that they are ideologically closer to China than to the United States or that they are vulnerable to the spread of liberal ideology, or both, the creation of a broad, U.S.-led balancing coalition against China is not nearly as likely as many realists predict.86 Philippine President Rodrigo Duterte, for instance, has consistently expressed greater ideological affinity for China than for the United States.87 It is therefore unsurprising that he has significantly weakened the Philippines’ alliance with the United States despite China's escalating material threat.88 Similar statements can be made about the likelihood of a U.S.-Russian alliance against China. Given Vladimir Putin's greater ideological affinity for China than for the United States and his obsession with the possibility of liberal revolution, a frenemy alliance with the United States to counter the material danger posed by China is highly unlikely as long as Putin's regime continues.89

In sum, analyzing only the effects of shared material threats will frequently provide a poor guide to predicting when ideological enemies are likely to form stable coalitions. To understand the likelihood of frenemy alliances, analysts must identify the conditions that make either the material forces that attract potential allies or the ideological forces that repel these states more salient to decisionmaking. Examining levels of regime vulnerability and the configurations of ideological distances among states will help leaders anticipate the nature of cross-ideological coalitions, both those that oppose their country and those that support it.

Notes

The author acknowledges with thanks comments from the anonymous reviewers, as well as participants in seminars sponsored by Brigham Young University, Cornell University, Dartmouth College, Johns Hopkins University, and Waseda University. He also is grateful to Clifford Bob, Timothy Crawford, Steven David, Robert Freedman, Gregory Gause, Harry Harding, David Lesch, Henry Nau, Chad Nelson, John Owen, Binnur Ozkececi-Taner, Michael Poznansky, Evan Resnick, Robert Ross, Jennie Schulze, Goktug Sonmez, Frederick Teiwes, Umut Uzer, Will Walldorf, Yafeng Xia, Keren Yarhi-Milo, and Andrew Yeo. These individuals commented on this article or the larger project from which it is drawn. The article is adapted from Frenemies: When Ideological Enemies Ally, by Mark L. Haas. Copyright © 2022 by Cornell University Press. Used by permission of the publisher, Cornell University Press

1.

As I detail below, one state poses a material threat to another based on a combination of superior military and economic power, geographical closeness, and displayed aggressiveness. Some prominent studies of alliances have mentioned the effects of ideological enmity and disputes as a potential barrier to alliance formation and durability. See Glenn H. Snyder, Alliance Politics (Ithaca, N.Y.: Cornell University Press, 1997), pp. 44, 46, 384 (fn. 3); George Liska, Nations in Alliance: The Limits of Interdependence (Baltimore, Md.: Johns Hopkins University Press, 1962), pp. 10, 16–18; and Stephen M. Walt, “Why Alliances Endure or Collapse,” Survival, Vol. 39, No. 1 (Spring 1997), pp. 162–163, doi.org/10.1080/00396339708442901. None, though, have given the subject systematic analysis in terms of detailing the conditions when ideological enmity is most and least likely to affect the creation and stability of alliances. The result, as Evan N. Resnick summarizes, is that “the dynamics of … alliances [among ideological enemies, or what Resnick labels ‘strange bedfellows'] have yet to be systematically explored by international relations scholars.” Resnick, “Strange Bedfellows: U.S. Bargaining Behavior with Allies of Convenience,” International Security, Vol. 35, No. 3 (Winter 2010/11), p. 144, doi.org/10.1162/ISEC_a_00026.

2.

Beginning in 1971, China and the United States did engage in a remarkable rapprochement that included repeated, high-level meetings and statements regarding the shared threat posed by the Soviet Union. Security cooperation between the two states remained, however, very limited. The result was that, until 1979, the two countries had, according to Robert S. Ross, only “the illusion of military relations.” Ross, Negotiating Cooperation: The United States and China, 1969–1989 (Stanford, Calif.: Stanford University Press, 1995), p. 88; see also p. 85.

3.

On the centrality of ideological calculations to these alliance outcomes, see the case study at the end of this article; Haas, Frenemies, chaps. 2–4; Frederick C. Teiwes and Warren Sun, The End of the Maoist Era: Chinese Politics During the Twilight of the Cultural Revolution, 1972–1976 (Armonk, N.Y.: M.E. Sharpe, 2007), pp. 130–133, 386, 427, 429, 433; Banu Eligur, The Mobilization of Political Islam in Turkey (Cambridge: Cambridge University Press, 2010), pp. 273–274; and Lisel Hintz, “‘Take it Outside!’ National Identity Contestation in the Foreign Policy Arena,” European Journal of International Relations, Vol. 22, No. 2 (June 2016), p. 351, doi.org/10.1177/1354066115588205.

4.

For details on these judgments about the likelihood of cross-ideological alliances, see Snyder, Alliance Politics, p. 152; D. Cameron Watt, “An Intelligence Surprise: The Failure of the Foreign Office to Anticipate the Nazi-Soviet Pact,” Intelligence and National Security, Vol. 4, No. 3 (July 1989), p. 529, doi.org/10.1080/02684528908432014; Andreas Etges, “All That Glitters Is Not Gold: The 1953 Coup against Mohammed Mossadegh in Iran,” in Isabelle Duyvesteyn, ed., Intelligence and Strategic Culture (New York: Routledge, 2013), p. 66; Vitaly Kozyrev, “Soviet Policy toward the United States and China, 1969–1979,” in William C. Kirby, Robert S. Ross, and Gong Li, eds., Normalization of U.S.-China Relations: An International History (Cambridge, Mass.: Harvard University Asia Center, 2005), p. 269; and David Johnston, “Threats and Responses: Washington; Top U.S. Officials Press Case Linking Iraq to Al Qaeda,” New York Times, February 12, 2003.

5.

For another analysis of the term “frenemy” that also connects states’ domestic politics to alliance policies, see Jeffrey W. Taliaferro, Defending Frenemies: Alliances, Politics, and Nuclear Nonproliferation in US Foreign Policy (Oxford: Oxford University Press, 2019). For related use of the term, see Merve Tahiroğlu and Behnam Ben Taleblu, “Turkey and Iran: The Best of Frenemies,” Turkish Policy Quarterly, Vol. 14, No. 1 (Spring 2015), pp. 123–134, http://turkishpolicy.com/Files/ArticlePDF/turkey-and-iran-the-best-of-frenemies-spring-2015-en.pdf; Meliha Benli Altunışık, “Bitter Frenemies: The Not-Quite-Alliance between Saudi Arabia and Turkey,” Foreign Affairs, May 15, 2012, https://www.foreignaffairs.com/articles/turkey/2012-05-15/bitter-frenemies; and Binnur Ozkececi-Taner, “From Allies to Frenemies and Inconvenient Partners: Image Theory and Turkish-Israeli Relations,” Perceptions, Vol. 17, No. 3 (Autumn 2012), pp. 105–129, https://dergipark.org.tr/en/pub/perception/issue/48983/624988.

6.

On the importance of ideological distances to international relations, see Mark L. Haas, The Ideological Origins of Great Power Politics, 1789–1989 (Ithaca, N.Y.: Cornell University Press, 2005); John M. Owen IV, The Clash of Ideas in World Politics: Transnational Networks, States, and Regime Change, 1510–2010 (Princeton, N.J.: Princeton University Press, 2010); and Henry R. Nau, Conservative Internationalism: Armed Diplomacy under Jefferson, Polk, Truman, and Reagan (Princeton, N.J.: Princeton University Press, 2015). Configurations of ideological distances develop a concept that I have introduced elsewhere and labeled “ideological polarity.” This variable examines the effects on foreign policies that are created by the number of distinct ideological groupings in a system. See Mark L. Haas, The Clash of Ideologies: Middle Eastern Politics and American Security (New York: Oxford University Press, 2012); Mark L. Haas, “Ideological Polarity and Balancing in Great Power Politics,” Security Studies, Vol. 23, No. 4 (2014), pp. 715–753, doi.org/10.1080/09636412.2014.964991; and F. Gregory Gause III, “Ideologies, Alignments, and Underbalancing in the New Middle East Cold War,” PS: Political Science and Politics, Vol. 50, No. 3 (July 2017), pp. 672–675, doi.org/10.1017/S1049096517000373.

7.

I use the terms “political leaders,” “elites,” and “policymakers” interchangeably. I do the same for “ideological enmity” and “ideological rivalry.”

8.

For additional studies of ideology and international relations, see Alexander L. George, “Ideology and International Relations: A Conceptual Analysis,” Jerusalem Journal of International Relations, Vol. 9, No. 1 (1987), pp. 1–21; Stephen M. Walt, Revolution and War (Ithaca, N.Y.: Cornell University Press, 1996); Henry R. Nau, At Home Abroad: Identity and Power in American Foreign Policy (Ithaca, N.Y.: Cornell University Press, 2002); Lawrence Rubin, Islam in the Balance: Ideational Threats in Arab Politics (Stanford, Calif.: Stanford University Press, 2014); Mingjiang Li, Mao's China and the Sino-Soviet Split: Ideological Dilemma (New York: Routledge, 2012); Nikola Mirilovic and Myunghee Kim, “Ideology and Threat Perceptions: American Public Opinion toward China and Iran,” Political Studies, Vol. 65, No. 1 (March 2017), pp. 179–98, doi.org/10.1177/0032321715614850; Peter Hays Gries, The Politics of American Foreign Policy: How Ideology Divides Liberals and Conservatives over Foreign Affairs (Stanford, Calif.: Stanford University Press, 2014); Nigel Gould-Davies, “Rethinking the Role of Ideology in International Politics during the Cold War,” Journal of Cold War Studies, Vol. 1, No. 1 (Winter 1999), pp. 90–109, doi.org/10.1162/15203970152521908; Mark Kramer, “Ideology and the Cold War,” Review of International Studies, Vol. 25, No. 4 (October 1999), pp. 539–576, https://www.jstor.org/stable/20097622; and David C. Engerman, “Ideology and the Origins of the Cold War, 1917–1962,” in Melvyn P. Leffler and Odd Arne Westad, eds., The Cambridge History of the Cold War, Vol. 1: Origins (Cambridge: Cambridge University Press, 2010), pp. 20–43.

9.

On these points, see Walt, The Origins of Alliances, pp. 35–36 (quotation); and Gause, “Ideologies, Alignments, and Underbalancing in the New Middle East Cold War,” p. 673.

10.

On these points, see Haas, Ideological Origins of Great Power Politics, pp. 5–18; Owen, Clash of Ideas, pp. 32–52; Walt, Revolution and War, pp. 32–45; and Nau, At Home Abroad, pp. 21–32.

11.

Charles A. Duelfer and Stephen Benedict Dyson, “Chronic Misperception and International Conflict: The U.S.-Iraq Experience,” International Security, Vol. 36, No. 1 (Summer 2011), p. 96, doi.org/10.1162/ISEC_a_00045.

12.

Dwight D. Eisenhower, “Address at the Iowa State Fair at Des Moines,” August 30, 1954, American Presidency Project, https://www.presidency.ucsb.edu/documents/address-the-iowa-state-fair-des-moines.

13.

Quoted in Karim Sadjadpour, Reading Khamenei: The World View of Iran's Most Powerful Leader (Washington, D.C.: Carnegie Endowment for International Peace, 2008), p. 15.

14.

Quoted in Henry L. Roberts, “Maxim Litvinov,” in Gordon A. Craig and Felix Gilbert, eds., The Diplomats, 1919–1939 (Princeton, N.J.: Princeton University Press, 1953), p. 366.

15.

On the preceding, see John J. Mearsheimer, The Tragedy of Great Power Politics (New York: W.W. Norton, 2001), pp. 48, 156, 269; and Walt, Origins of Alliances, pp. 17, 21–26, 32, 37–38.

16.

For others who have used this approach, see Walt, Origins of Alliances; and Randall L. Schweller, “Bandwagoning for Profit: Bringing the Revisionist State Back In,” International Security, Vol. 19, No. 1 (Summer 1994), p. 83, doi.org/10.2307/2539149.

17.

For others who have examined the effects of regime vulnerability on states’ foreign policies, see Li, Mao's China and the Sino-Soviet Split; Owen, Clash of Ideas; Rubin, Islam in the Balance; and Randall L. Schweller, Unanswered Threats: Political Constraints on the Balance of Power (Princeton, N.J.: Princeton University Press, 2006).

18.

For related analysis, see Chad E. Nelson, Revolutionary Contagion and International Politics (Oxford: Oxford University Press, forthcoming); and Walt, Revolution and War, pp. 39–41.

19.

For related analysis, see Walt, “Why Alliances Endure or Collapse,” pp. 161–162.

20.

These last dynamics go far in explaining Turkey's Islamist leaders’ actions over the course of 2009 and 2010 that ended their country's frenemy alliance with Israel. See Eligur, The Mobilization of Political Islam in Turkey, pp. 273–274, 282.

21.

Robert Loeffel, The Fifth Column in World War II: Suspected Subversives in the Pacific War and Australia (New York: Palgrave Macmillan, 2015).

22.

Alex Braithwaite, “Resisting Infection: How State Capacity Conditions Conflict Contagion,” Journal of Peace Research, Vol. 47, No. 3 (May 2010), pp. 311–319, doi.org/10.1177/0022343310362164.

23.

See Lester R. Kurtz and Lee A. Smithey, eds., The Paradox of Repression and Nonviolent Movements (Syracuse, N.Y.: Syracuse University Press, 2018).

24.

For similar analysis, see Wade Jacoby, “Inspiration, Coalition, and Substitution: External Influences on Postcommunist Transformations,” World Politics, Vol. 58, No. 4 (July 2006), pp. 625–626, 629, 645, doi.org/10.1353/wp.2007.0010.

25.

Ibid., p. 645, also pp. 625–626, 629; and Rubin, Islam in the Balance, p. 23.

26.

Li, Mao's China and the Sino-Soviet Split, p. 117. The Sino-Soviet case is a prime example of how a divisive ideology and competition over leadership of the transnational ideological group can result in intense ideological enmity among co-ideologies.

27.

James M. Lutz, “The Diffusion of Political Phenomena in Sub-Saharan Africa,” Journal of Political & Military Sociology, Vol. 17, No. 1 (Spring 1989), pp. 93–114, https://www.jstor.org/stable/45293500; Daniel Brinks and Michael Coppedge, “Diffusion Is No Illusion: Neighbor Emulation in the Third Wave of Democracy,” Comparative Political Studies, Vol. 39, No. 4 (May 2006), pp. 463–489, doi.org/10.1177/0010414005276666; Samuel P. Huntington, The Third Wave: Democratization in the Late Twentieth Century (Norman: University of Oklahoma Press, 1991), pp. 100–104; and Owen, Clash of Ideas, pp. 38–42.

28.

Henry E. Hale, “Regime Change Cascades: What We Have Learned from the 1848 Revolutions to the 2011 Arab Uprisings,” Annual Review of Political Science, Vol. 16, No. 1 (May 2013), pp. 331–353, doi.org/10.1146/annurev-polisci-032211-212204.

29.

For an example of these dynamics, see M.E. Sarotte, “China's Fear of Contagion: Tiananmen Square and the Power of the European Example,” International Security, Vol. 37, No. 2 (Fall 2012), pp. 156–182, doi.org/10.1162/ISEC_a_00101.

30.

Colin J. Beck, “The World-Cultural Origins of Revolutionary Waves: Five Centuries of European Contention,” Social Science History, Vol. 35, No. 2 (Summer 2011), pp. 182–183, doi.org/10.1017/S0145553200011482.

31.

Owen, Clash of Ideas, pp. 5, 38–42.

32.

Shahram Akbarzadeh, “Uzbekistan and the United States: Friends or Foes?” Middle East Policy Council, Vol. 14, No. 1 (Spring 2007), pp. 107–116, doi.org/10.1111/j.1475-4967.2007.00288.x.

33.

Ibid., p. 114.

34.

On policies of ideological exportation, see John M. Owen IV, “The Foreign Imposition of Domestic Institutions,” International Organization, Vol. 56, No. 2 (Spring 2002), pp. 375–409, doi.org/10.1162/002081802320005513; Suzanne Werner, “Absolute and Limited War: The Possibility of Foreign-Imposed Regime Change,” International Interactions, Vol. 22, No. 1 (July 1996), pp. 67–88, doi.org/10.1080/03050629608434880; and Nigel Lo, Barry Hashimoto, and Dan Reiter, “Ensuring Peace: Foreign-Imposed Regime Change and Postwar Peace Duration, 1914–2001,” International Organization, Vol. 62, No. 4 (October 2008), pp. 717–736, doi.org/10.1017/S0020818308080259.

35.

Quotations from Julian Gewirtz, Unlikely Partners: Chinese Reformers, Western Economists, and the Making of Global China (Cambridge, Mass.: Harvard University Press, 2017), p. 3.

36.

Ross, Negotiating Cooperation, p. 88 (first quotation); and Raymond L. Garthoff, Détente and Confrontation: American-Soviet Relations from Nixon to Reagan (Washington, D.C.: Brookings Institution Press, 1995), p. 690 (second quotation).

37.

Quoted in Sergey Radchenko, “Sino-Soviet Relations in the 1970s and IR Theory,” in Gilbert Rozman, ed., Misunderstanding Asia: International Relations Theory and Asian Studies over Half a Century (New York: Palgrave Macmillan, 2015), p. 58.

38.

Nelson, Revolutionary Contagion and International Politics, chap. 6.

39.

By designating an initiating state, I am simply identifying a country that is deciding whether to commit to an alliance. The initiating state is not necessarily the prime mover for the coalition. The point is to examine alliance policies from this country's perspective.

40.

William I. Shorrock, From Ally to Enemy: The Enigma of Fascist Italy in French Diplomacy, 1920–1940 (Kent, Ohio: Kent State University Press, 1988), pp. 104, 118, 168.

41.

Quotations from Robert H. Whealey, “Mussolini's Ideological Diplomacy: An Unpublished Document,” Journal of Modern History, Vol. 39, No. 4 (December 1967), pp. 435–436, https://www.journals.uchicago.edu/doi/10.1086/240124.

42.

On the need for elite consensus regarding the nature of international threats for effective balancing, see Randall L. Schweller, “Unanswered Threats: A Neoclassical Realist Theory of Underbalancing,” International Security, Vol. 29, No. 2 (Fall 2004), p. 171, doi.org/10.1162/0162288042879913.

43.

Zara Steiner, The Triumph of the Dark: European International History, 1933–1939 (Oxford: Oxford University Press, 2011), pp. 437–438. On the centrality of ideological differences to Britain's relations with the Soviet Union in the 1930s, see Louise Grace Shaw, The British Political Elite and the Soviet Union, 1937–1939 (London: Frank Cass, 2003).

44.

“Quotations from Report by Four Chinese Marshals—Chen Yi, Ye Jianying, Xu Xiangqian, and Nie Rongzhen—to the Central Committee, ‘A Preliminary Evaluation of the War Situation,'” July 11, 1969, doc. no. 9, Cold War International History Project Bulletin, No. 11 (Winter 1998), p. 166, https://www.wilsoncenter.org/sites/default/files/media/documents/publication/CWIHP_Bulletin_11.pdf.

45.

“The Great Victory of Chairman Mao's Revolutionary Diplomatic Line,” Confidential Reference Material on Education on the Situation, no. 43, for Distribution to Companies and Above, April 4, 1973, quoted in Garthoff, Détente and Confrontation, p. 240.

46.

Rachel Bronson, Thicker Than Oil: America's Uneasy Partnership with Saudi Arabia (Oxford: Oxford University Press, 2006), p. 26.

47.

On strategies to wedge apart allies or potential allies, see Timothy W. Crawford, The Power To Divide: Wedge Strategies in Great Power Competition (Ithaca, N.Y.: Cornell University Press, 2021); and Yasuhiro Izumikawa, “To Coerce or Reward? Theorizing Wedge Strategies in Alliance Politics,” Security Studies, Vol. 22, No. 3 (2013), pp. 498–531, doi.org/10.1080/09636412.2013.816121.

48.

M.B. Hayne, The French Foreign Office and the Origins of the First World War, 1898–1914 (Oxford: Clarendon, 1993), pp. 108, 112, 245.

49.

Georges Michon, The Franco-Russian Alliance, 1891–1917 (New York: Macmillan, 1929), p. 105, also pp. 102, 203–204, 260.

50.

I have already discussed how the ideological configuration of ideological outsider facilitated France's frenemy alliance with Russia. Low levels of regime vulnerability worked for the same outcome. The most powerful French leaders did not believe that allying with czarist Russia threatened their domestic interests. Ibid., pp. 73–75, 81, 86–87. This made the domestic costs of committing to the frenemy alliance small. The values of the argument's independent variables for French leaders in relation to Russia before World War I were thus the opposite of those that existed for France in relation to the Soviet Union in the 1930s. This variation goes far in explaining the opposing alliance policies toward Russia/the Soviet Union before the world wars.

51.

Cheng-Chwee Kuik, “How Do Weaker States Hedge? Unpacking ASEAN States’ Alliance Behavior towards China,” Journal of Contemporary China, Vol. 25, No. 100 (2016), p. 504, doi.org/10.1080/10670564.2015.1132714.

52.

See Paul M. Kennedy, The Rise of the Anglo-German Antagonism, 1860–1914 (Amherst, N.Y.: Humanity, 1988), pp. 157–158, 151–165, 188–201. On Germany's low level of regime vulnerability, see ibid., pp. 186–188; and Owen, Clash of Ideas, pp. 154–156.

53.

Peter Jackson, France and the Nazi Menace: Intelligence and Policy Making, 1933–1939 (Oxford: Oxford University Press, 2000), p. 207.

54.

One, or the other, or both groups led all governments but one in these years (excluding governments that lasted less than one month). This exception occurred from June 1936 to June 1937, when the socialist Léon Blum was prime minister.

55.

William Evans Scott, Alliance against Hitler: The Origins of the Franco-Soviet Pact (Durham, N.C.: Duke University Press, 1962), pp. 135, 137–138, 189–190, 193–197, 219, 235–236, 239, 244–246, 251–252, 254.

56.

Quoted in ibid., p. 239.

57.

Quoted in ibid., p. 247.

58.

Ibid., p. 264.

59.

Michael Jabara Carley, “End of the ‘Low, Dishonest Decade': Failure of the Anglo-Franco-Soviet Alliance in 1939,” Europe-Asia Studies, Vol. 45, No. 2 (1993), p. 306, https://www.jstor.org/stable/152863.

60.

Quoted in Michael Jabara Carley, “Prelude to Defeat: Franco-Soviet Relations, 1919–39,” Historical Reflections, Vol. 22, No. 1 (Winter 1996), pp. 182–183, https://www.jstor.org/stable/41299054. See also Michael Jabara Carley, “A Soviet Eye on France from the rue de Grenelle in Paris, 1924–1940,” Diplomacy and Statecraft, Vol. 17, No. 2 (2006), pp. 322–323, doi.org/10.1080/09592290600695292.

61.

Anthony Adamthwaite, France and the Coming of the Second World War, 1936–1939 (London: Frank Cass, 1977), pp. 312, 327, 335–336, 354.

62.

Steiner, The Triumph of the Dark, p. 906; see also Carley, “End of the ‘Low, Dishonest Decade,'” pp. 326, 328, 332.

63.

Edward Shorter and Charles Tilly, Strikes in France, 1830–1968 (Cambridge: Cambridge University Press, 1974), p. 314; and Scott, Alliance against Hitler, p. 156.

64.

Scott, Alliance against Hitler, p. 262.

65.

Peter J. Larmour, The French Radical Party in the 1930's (Stanford, Calif.: Stanford University Press, 1964), pp. 152, 172–176.

66.

Scott, Alliance against Hitler, p. 264; and Carley, “A Soviet Eye on France from the rue de Grenelle in Paris,” pp. 315, 319.

67.

Quoted in Charles A. Micaud, The French Right and Nazi Germany, 1933–1939 (Durham, N.C.: Duke University Press, 1943), p. 80, also p. 111; and William D. Irvine, French Conservatism in Crisis: The Republican Federation of France in the 1930s (Baton Rouge: Louisiana State University Press, 1979), p. 167.

68.

Micaud, The French Right and Nazi Germany, pp. 83–110; and Irvine, French Conservatism in Crisis, pp. 83–84, 88.

69.

David A.L. Levy, “The French Popular Front, 1936–37,” in Helen Graham and Paul Preston, eds., The Popular Front in Europe (London: Macmillan, 1987), pp. 75–78.

70.

Jackson, France and the Nazi Menace, p. 179; Larmour, The French Radical Party in the 1930s, pp. 177, 213, 218, 245, 253; and Carley, “End of the ‘Low, Dishonest Decade,'” p. 307.

71.

Quoted in Carley, “A Soviet Eye on France from the rue de Grenelle in Paris,” p. 323.

72.

Nicole Jordan, The Popular Front and Central Europe: The Dilemmas of French Impotence, 1918–1940 (New York: Cambridge University Press, 1992), p. 228. The other two main sources of regime vulnerability—belief that the Soviet Union was endeavoring to export its ideology to France and the anticipated impact of demonstration effects created by potential revolution in other countries (chiefly Spain)—significantly amplified Radicals’ and conservatives’ fears of regime vulnerability created by the communist fifth column, thereby making the costs of an alliance with the Soviet Union even higher. For details, see John E. Dreifort, “The French Popular Front and the Franco-Soviet Pact, 1936–37: A Dilemma in Foreign Policy,” Journal of Contemporary History, Vol. 11, No. 2/3 (July 1976), p. 226, doi.org/10.1177/002200947601100212; and Anthony Adamthwaite, Grandeur and Misery: France's Bid for Power in Europe, 1914–1940 (London: Arnold, 1995), p. 207. Space constraints prevent examination of these additional factors.

73.

Quoted in Jordan, The Popular Front and Central Europe, p. 228.

74.

See, for example, Andrew Rothstein, The Munich Conspiracy (London: Lawrence and Wishart, 1958), pp. 23, 103; and Micaud, The French Right and Nazi Germany, p. 99.

75.

Quoted in Joel Colton, Léon Blum: Humanist in Politics (New York: Alfred A. Knopf, 1966), p. 199.

76.

Ibid., p. 210, also p. 211.

77.

Jordan, The Popular Front and Central Europe, pp. 263, 272, 313; Jackson, France and the Nazi Menace, pp. 236–237; and Colton, Léon Blum, p. 211.

78.

Nathanael Greene, Crisis and Decline: The French Socialist Party in the Popular Front Era (Ithaca, N.Y.: Cornell University Press, 1969), pp. 23–25.

79.

Ibid., pp. 28–29.

80.

French socialists’ fears of regime vulnerability in relation to communism also were relatively low. Indeed, these elites viewed French communists as important allies against the rise of French fascism, hence their domestic alliance with communists in the Popular Front. See ibid., pp. 9–18.

81.

Thomas J. Christensen and Jack Snyder, “Chain Gangs and Passed Bucks: Predicting Alliance Patterns in Multipolarity,” International Organization, Vol. 44, No. 2 (Spring 1990), pp. 159–165, doi.org/10.1017/S0020818300035232; Kenneth N. Waltz, Theory of International Politics (New York: McGraw-Hill, 1979), p. 165; and Mearsheimer, The Tragedy of Great Power Politics, pp. 308–313.

82.

Barry R. Posen, The Sources of Military Doctrine: France, Britain, and Germany between the World Wars (Ithaca, N.Y.: Cornell University Press, 1984), p. 123, also p. 125; Robert J. Young, In Command of France: French Foreign Policy and Military Planning, 1933–1940 (Cambridge, Mass.: Harvard University Press, 1978), p. 58; and Jean-Baptiste Duroselle, France and the Nazi Threat: The Collapse of French Diplomacy, 1932–1939 (New York: Enigma, 2004), pp. 238, 269.

83.

See, for example, Steven R. David, Choosing Sides: Alignment and Realignment in the Third World (Baltimore, Md.: Johns Hopkins University Press, 1991); and F. Gregory Gause III, “Balancing What? Threat Perception and Alliance Choice in the Gulf,” Security Studies, Vol. 13, No. 2 (2003), pp. 273–305, doi.org/10.1080/09636410490521271.

84.

On the tendency for international relations of some periods to be highly ideological and others not, as well as a related explanation for the variation, see Owen, Clash of Ideas.

85.

Gause used an early version of the argument to explain why a Saudi-Turkish-Israeli alliance against Iran has yet to form. Gause, “Ideologies, Alignments, and Underbalancing in the New Middle East Cold War”; and Haas, “Ideological Polarity and Balancing in Great Power Politics.”

86.

On this prediction, see John J. Mearsheimer, “Can China Rise Peacefully?” National Interest, October 25, 2014, https://nationalinterest.org/commentary/can-china-rise-peacefully-10204?page=0%2C4.

87.

While Duterte has announced that he has “realigned myself in [China's] ideological flow,” he also condemned the United States as imperialist and racist. Ben Blanchard, “Duterte Aligns Philippines with China, Says U.S. Has Lost,” Reuters, October 20, 2016, https://www.reuters.com/article/us-china-philippines/duterte-aligns-philippines-with-china-says-u-s-has-lost-idUSKCN1-2K0AS; and Prashanth Parameswaran, “Why the Philippines’ Rodrigo Duterte Hates America,” Diplomat, November 1, 2016, https://thediplomat.com/2016/11/why-the-philippines-rodrigo-duterte-hates-america/.

88.

Richard Javad Heydarian, “Tragedy of Small Power Politics: Duterte and the Shifting Sands of Philippine Foreign Policy,” Asian Security, Vol. 13, No. 3 (2017), p. 233, doi.org/10.1080/14799855.2017.1354569.

89.

On these points, see Brian D. Taylor, The Code of Putinism (New York: Oxford University Press, 2018), pp. 6, 11, 22, 15, 17–18, 20, 25, 27, 35, 171, 176–179, 182, 184–185, 188, 192; and Michael McFaul, “Putin, Putinism, and the Domestic Determinants of Russian Foreign Policy,” International Security, Vol. 45, No. 2 (Fall 2020), pp. 95–139, doi.org/10.1162/isec_a_00390.