Beyond confrontation: A typology of proxy wars between rival powers

Abstract

The study of how rival powers interact when they engage in proxy wars has generated many important insights, but the field remains fragmented. This article integrates these insights by offering a typological framework focused on interaction patterns among rival powers. The framework rests on two key choices rival powers face when conducting proxy interventions in third-party conflicts. The first is whether to support the same or opposing local warring parties; the second is whether to coordinate their in-theater activities. Combining these two dimensions yields four distinct interaction patterns, which we call (1) confrontation, (2) managed rivalry, (3) competitive influence-seeking, and (4) tacit alignment. The article unpacks these categories, specifies the mechanisms through which they emerge, and illustrates the framework’s analytical value through a series of empirical vignettes drawn from both the Cold War and the contemporary era. The framework enables structured within-case and between-case comparisons as well as large-n analyses aimed at uncovering recurring conditions associated with distinct forms of interaction between rival powers.

Introduction

Proxy wars have been the subject of a burgeoning literature in the last 15 years (Berman and Lake, 2019; Farasoo, 2021; Groh, 2019; Hughes, 2012; Moghadam et al., 2023; Mumford, 2013; Rauta, 2018; San-Akca, 2016). This article contributes to that body of scholarship by examining a key aspect: how rival powers engage one another in proxy wars. Understanding this dynamic is important. After all, proxy wars involving rival powers are a regular feature of international politics, especially during periods of intensified great power competition. The paradigmatic example is the Cold War during which the United States and the Soviet Union relied heavily on proxies to intervene in conflicts across the globe (for overviews, see Falkov, 2023; Westad, 2005). Hence, as we are now drifting toward a world of competing power centers, many observers have pointed to the renewed importance of wars conducted through local surrogates. As one analyst put it, “great-power competition has always gone hand-in-hand with proxy wars of various kinds; there is little reason to expect this time to be different” (Winter-Levy, 2022; see also Groh, 2019: 46; Mumford, 2013: 90–96, 110–112). Indeed, some of the most prominent and deadliest civil wars in the past 10 years have been subject to involvements by rival powers. Examples include the conflicts in Syria, Libya, Sudan, Yemen, and Ukraine before Russia’s large-scale invasion in February 2022 (Byman, 2023).

Drawing on the literature on strategic rivalries, we define rival powers as states engaged in “long-term conflict-ridden relations” marked by “perceptions of mutual hostility [and] mistrust” (Maoz and San-Akca, 2012: 720). In such relationships, the threat and use of force are ever-present possibilities (Goertz and Diehl, 1994: 154). One might assume, therefore, that rival powers interfering in another country’s internal conflict will invariably assist opposing warring parties and act in a highly confrontational manner. Even a cursory look at the empirical record shows that this is not the case, however. Consider a few examples.

Despite an increasingly adversarial relationship between the United States and Russia, the two continued to cooperate with each other in Afghanistan at least until 2014. They coordinated the supply of weapons and equipment to the Northern Alliance fighting the Taliban, shared intelligence, and even conducted joint anti-narcotic raids (Schwirtz, 2010). In the Syrian civil war, meanwhile, Russia and the United States supported opposing sides. Nevertheless, they worked assiduously to prevent direct clashes between themselves and their respective proxy partners. To this end, they established de-confliction protocols for both air and land operations (Weiss and Ng, 2019). This, in turn, stands in contrast to conflict dynamics that played out in Mali. In 2020–2022, both a French-led coalition of Western states and Russia backed the central government fighting a jihadist insurgency, but there was no cooperation between them. On the contrary, France and Russia competed intensely against one another to be the primary ally and patron of Mali’s government (Le Roux, 2022). In short, rival powers can interact in strikingly different ways when involved in proxy wars.

Conflict researchers know this, of course. Thus far, however, there exists no unified framework that captures this variation. As shown below, work on the employment of proxy warfare by rival powers is scattered across related but hitherto separate research clusters. Each cluster has produced many relevant and important findings, but often under different labels. Moreover, each cluster tends to highlight one or several types of rival power interactions, while neglecting others. As a result, the scholarship remains fragmented. ¹ The goal of this article is to introduce a typological framework that encompasses the whole range of rival power interactions in proxy wars and to offer a shared conceptual vocabulary that bridges different research clusters. Our study thus picks up on Mumford’s call for “greater terminological clarity and theoretical integration” in proxy war scholarship (Mumford, 2021: 2055).

In brief, the proposed framework identifies four principal interaction patterns between rival powers engaged in proxy warfare: (1) confrontation, (2) managed rivalry, (3) competitive influence-seeking, and (4) tacit alignment. Confrontation refers to cases where rival powers back opposing sides in another country’s civil war and act in a highly antagonistic manner. Managed rivalry describes situations in which rival powers back opposing sides but simultaneously coordinate their in-theater activities. Competitive influence-seeking describes a dynamic in which rival powers become involved on the same side in another country’s civil war but they do not collaborate with each other; instead, they seek to outbid one another to be the primary patron of the local actor they support. Finally, tacit alignment refers to cases in which rival powers back the same actor in another country’s internal conflict and cooperate, at least to some degree. These four categories capture, in our view, much of the key variation in how rival powers interact in proxy wars.

The remainder of this articles develops the framework in detail. Toward that end, we proceed as follows. The first section reviews the existing proxy war literature and shows that, despite much progress, there is no unified framework that captures the full range of rival power interactions. The second section introduces our four-pronged typology, lays out the basic building blocks and assumptions that underpin it, and elaborates on the mechanisms through which different interaction patterns emerge. It also illustrates the typology’s analytical traction and utility through a series of empirical vignettes drawn from both the Cold War and the contemporary era. The third section offers a brief summary and, based on this, identifies promising avenues for future research.

Situating this study in the existing literature

Much of the existing proxy war literature focuses on the relationship between an outside power and its local partner, often by employing insights from principal-agent theory. In particular, many studies seek to explore the difficulties of an outside power (principal) to enjoy reasonable levels of oversight and control over a local partner (agent) given divergent preferences and asymmetric information (Berman and Lake, 2019; Karlén and Rauta, 2023; Popovic, 2017; Salehyan, 2010; Salehyan et al., 2011). This emphasis on principal-agent relationships is understandable. After all, it is a defining feature of proxy warfare. At the same time, these studies show little concern for how the interplay between two or more outside powers affects local conflict dynamics. In other words, while numerous works analyze the vertical relationship between a sponsor and its proxy, far less attention has been paid to the horizontal relationship between two rival powers intervening in the same conflict. As Hoffman and Orner (2023: 428) observe, “The literature is often silent on the principal’s ultimate competitor, the other great power.”

That being said, the point should not be overstated. Several strands of research offer pertinent insights into how rival powers interact in proxy wars. One such strand is found in the literature on third-party interventions—both direct and indirect—in civil wars. Work in this vein has established that the involvement of one outside power in a civil war increases the likelihood that another will become engaged as well (Anderson, 2019; Corbetta and Dixon, 2005; Findley and Teo, 2006; Maoz and San-Akca, 2012; Mitton, 2023; Toukan, 2019). A range of labels have been used to describe this dynamic, including “balanced interventions” (Balch-Lindsay et al., 2008), “dual-sided interventions” (Hironaka, 2008), “competitive interventions” (Anderson, 2019), and “balancing interventions” (Regan, 2002). ² Many scholars argue that this recurring pattern is driven by power-political considerations: rival powers seek to prevent each other from gaining influence, which could upset the global and/or regional balance of power (Anderson, 2019: 692–697; Mitton, 2023; Toukan, 2019). This captures something important about proxy wars, and we return to this point below. Yet, as noted in the introduction, it represents only one possible form of interaction between rivals. In some cases, even outright adversaries collaborate in a local conflict; in other cases, they vie for the loyalty of the same actor. The notion of competitive interventions, while useful, is therefore too narrow to capture the full range of rival power interactions in proxy wars.

A second cluster of relevant research emerges from the literature on escalation management. In a highly nuanced study, Wilkenfeld and Murauskaite (2023) demonstrate that major powers often turn to proxy warfare—rather than direct interventions—because of escalation concerns. ³ Similarly, Carson (2018) shows that major powers frequently pursue covert interventions, often by bolstering local forces in a third-party conflict, even though such activities are usually detected by rivals. The reason is that covert interventions signal a measure of restraint. Moreover, it allows decision-makers of rival powers to collude on the “backstage” while shielding themselves from hawkish domestic pressures (see also Cormac and Aldrich, 2018; Poznansky, 2022). The broader point here is that escalation concerns tend to impact rival powers’ decision-making calculus about how to engage one another in proxy war settings.

A third strand of literature examines proxy wars through the prism of the re-emerging—and by most accounts intensifying—great power competition. Many works in this genre suggest that proxy warfare is, or is likely to become, an important feature of international politics (Hoffman and Orner, 2023; Jenne and Siroky, 2023; Winter-Levy, 2022; Yeisley, 2011). In extension, several scholars have presented typologies of proxy wars involving great power rivals. For instance, Jenne and Siroky (2023) distinguish between defensive proxy interventions, where a great power backs an embattled government in another country; offensive proxy operations, where a great power supports an insurgency to overthrow another country’s government; and symmetrical proxy interventions, where two great powers support opposing sides. Another example is the study by Hoffman and Orner (2023), which outlines six distinct proxy intervention strategies of great powers (bleed, block I, balance, bait, block II, and blemish). Both typologies are useful but incomplete, as they share an inbuilt assumption—challenged here—that rival powers invariably support opposing sides in proxy wars.

A notable exception is the study by Tierney (2021). Focusing on US–China relations, he specifically argues that major powers may be “sparring in one civil war while steering clear of each other or even cooperating in another internal conflict” (Tierney, 2021: 50). This is a valuable insight, which we incorporate into our framework. Tierney, however, does not explain why and when rival powers tend to cooperate. Nor does his study consider the practice of rival powers engaging in what we call competitive influence-seeking; that is, cases in which they vie for the favor of the same local actor.

Finally, there exists a rich reservoir of case study work on specific proxy wars—often conducted by area specialists—that provide highly detailed accounts of why, when, and how external powers intervene through local surrogates (see Bergen et al., 2022; Hinnebusch and Saouli, 2019; Prunier, 2004). These studies show, among other things, that even entrenched enemies sometimes coordinate their military activities in local conflicts. By their own design, however, such studies seek to provide fine-grained narratives of specific conflicts rather than build generalizable frameworks that can be applied to a large universe of cases.

In sum, different strands of research have established, in broad terms, that there exists a great deal of variation in the way—and with which effects—rival powers engage in proxy warfare. What we lack, however, is a unifying framework that captures this variation. As a result, we know little about the conditions under which rivals will pursue cooperative or competitive strategies when sponsoring third parties in a local conflict; nor is there much agreement on the labeling of these various dynamics. This fragmented state of the literature provides the jumping-off point for the typological framework we develop in the next section.

Interaction patterns between rival powers in proxy wars: a typological framework

Although numerous definitions of the term “proxy war” exist, one of the most recent and widely cited comes from Rauta. According to Rauta (2018: 457), a proxy war can be understood as,

a violent armed interaction resulting from the polarization of competing political goals between two organized parties, a Beneficiary and a Target, in which at least one party engages the other indirectly in sustained collective violence through a third party, the Proxy.

Simply put, the focus is on one-sided delegation of war by an external power—or Beneficiary, to use Rauta’s terminology—to an armed group in a foreign country fighting another local actor. This definition serves the purposes of our article well, as it carries the advantage of avoiding a priori assumptions about the number of external powers or their interactions, while still capturing the essential element of delegating war to a local actor. ⁴

An added advantage is that Rauta’s formulation subsumes Karl Deutsch’s classical understanding of proxy wars. According to Deutsch, a proxy war is best understood as “an international conflict between two foreign powers, fought out on the soil of a third country; disguised as conflict over an internal issue of that country” (Deutsch, 1964: 102). The emphasis here is on competing powers backing opposing sides in a local conflict as an extension of their broader rivalry (see also Jenne and Siroky, 2023: 439). In an era of growing great power competition, Deutsch’s definition may appear particularly relevant. Yet it is also overly restrictive, as it presumes that rival powers inevitably will confront one another through local surrogates. In practice, rivals may support proxies in the same conflict theater without necessarily clashing. Thus, we adopt Rauta’s broader definition, which better accommodates the full spectrum of rival power behavior in proxy war settings.

As indicated earlier, our framework suggests that one can distinguish four different interaction patterns. This typology is based on two basic choices that rivals face when they engage in proxy warfare. The first choice is which actor to support in another country’s internal armed conflict. For our purposes, it is less important whether rival powers throw their weight behind the incumbent government or an insurgency group. What matters is whether rival powers decide to back the same local actor or opposing ones. The second choice concerns the question whether rival powers, despite their overall antagonistic relationship, are willing to manage their engagement in the local conflict theater. By this, we mean that rival powers coordinate their involvement through the establishment of specific communication channels, agreements on distinct geographic areas of operation, or otherwise work together in the war zone (more on which below). Combining these two dimensions yields four distinct interaction patterns between rival powers, as visualized in Figure 1.

Figure 1.

Interaction patterns between rival powers in proxy wars.

In the upper-left cell, rival powers are unwilling to manage their engagement in the local conflict theater and support opposing warring parties. This typically produces highly confrontational dynamics. In the upper-right cell, rival powers also support opposing warring parties in a local conflict but they are willing and able to manage their in-theater engagement. As a result, one can observe what we call managed rivalries. In the lower-left cell, rival powers back the same local actor but are unwilling to manage their in-theater operations. This results in competitive influence-seeking, which is to say that one power seeks to outcompete the other to become the local actor’s principal patron. Finally, in some cases, even highly antagonistic powers decide to support the same local actor and to coordinate their in-theater activities. This leads to a form of tacit alignment between the rivals, which is the lower-right cell in Figure 1. In short, our argument is that rival power interactions in proxy wars can take one of four forms. Below, we further unpack each type in greater detail. Before proceeding, however, several clarifications are in order.

First, as should be clear from the previous discussion, the proposed typology hones in on interaction patterns among rivals interfering in another country’s armed conflict. This is not to imply that local warring parties are obedient puppets willfully carrying out the wishes of external powers. Local parties obviously follow their own agendas and often exercise significant agency in choosing their sponsors. Put simply, being a proxy is not the same as being a pawn (Chinchilla et al., 2021: 2073; San-Akca, 2016: 105–137). That being said, proxy relations are often asymmetric insofar as external sponsors typically command greater military and economic resources than their local partners (Groh, 2019: 119; Dunér, 1981; Moghadam and Wyss, 2018). ⁵

Second, our focus on strategic rivals implies that joint proxy interventions by aligned or allied powers fall outside the bounds of our framework. Note also that while the proposed typology may be extended to non-state actors as sponsors of proxies (Moghadam and Wyss, 2020), the focus in this study is on states as relevant external actors.

Third, the type of support provided by rival powers to a proxy can range from offering training, equipment, and weapons to the deployment of combat troops (Byman et al., 2001: 83–100; Hoffman and Orner, 2023; Mumford, 2013: 61–69). Some scholars question whether the deployment of combat troops into a conflict theater can be considered a form of proxy warfare, or whether it crosses the threshold and becomes a full-blown war (for contending views, see Groh, 2019: 29; Mumford, 2013: 15; Stark, 2021). Our position is that outside powers deploying troops alongside a local conflict party can be understood as a form of proxy warfare, as long as it is done on behalf of a local actor and that actor significantly contributes to the fighting on the ground (Watts et al., 2023: 4).

With these considerations in mind, we now describe the different types of interaction between rival powers and identify some of the mechanisms through which they are likely to emerge. While we leave a detailed analysis of the causes for future research, we highlight in particular (but not exclusively) the importance of threat perceptions and risk assessments. For each type, we provide two empirical illustrations: one drawn from the Cold War and one from the contemporary era. ⁶ In pairing examples from different periods, we seek to demonstrate the framework’s broad applicability. For each example, we draw on a range of secondary sources along with publicly available primary material.

Confrontation

One set of proxy wars involving rival powers is marked by confrontational dynamics, to put it in the conceptual language of our framework. This means that two rivals back opposite sides in another country’s civil war and engage in highly adversarial behavior in the local conflict theater. Of course, there exist numerous possible motives for states to support a proxy in the first place, ranging from security concerns to ethno-religious sentiments to economic interests (Byman et al., 2001: 23–40; Mumford, 2013: 30–44; Salehyan et al., 2011: 712–715; Watts et al., 2023: 13–53). What is important for this article’s purposes is that once one external power becomes involved, its rival is likely to become involved as well.

There are at least two reasons for this: First, in eras of heightened strategic competition, zero-sum thinking comes to the fore. Rival powers fear that even a small number of states drawn into the orbit of the other could upset the existing balance of power, influence, and/or status. Hence, each sees the involvement of the other in third-party conflicts as a potential threat that needs to be countered (Anderson, 2019: 694–695; Mitton, 2023: 19–43; Toukan, 2019: 815–816). Second, the involvement of an external power in another country’s civil war presents an opportunity for the rival to respond in ways that prolong this war. The aim is to bog down the intervening power in a bloody quagmire by bolstering the capabilities of the local adversary (Grauer and Tierney, 2018: 271–272; Hoffman and Orner, 2023). In short, rival powers tend to engage in confrontational proxy wars either to prevent losses or to entangle the other in a costly conflict. Some cases display a combination of both. The result is a move-countermove dynamic whereby the involvement of one external power induces a response by the other, and so on. ⁷

An example from the Cold War is the Afghan civil war that started in 1978. The Soviet Union provided military and economic support to the country’s communist government, which faced an Islamist-inspired insurgency. The insurgents, for their part, received some aid from the United States, initially on a small scale (Gates, 1996: 146–147). Then, in late 1979, the Soviet Union invaded Afghanistan with 80,000 troops. As several archival-based studies have shown, one of the primary drivers behind this decision was Moscow’s fear that the Afghan leadership was on the verge of switching loyalties from the USSR toward the West (Gibbs, 2006; Westad, 1994).

For the United States, the Soviet invasion presented both a challenge and an opportunity. On one hand, US policymakers were concerned that the Soviets might use Afghanistan as a springboard to extend their influence further south. As President Jimmy Carter in his memoirs noted, “a successful takeover of Afghanistan” would have given “the Soviets a deep penetration between Iran and Pakistan and pose a threat to the rich oil fields of the Persian Gulf area” (Carter, 1982: 481). On the other hand, government officials in Washington saw the invasion also as an opportunity to get its Cold War rival bogged down in an unwinnable and costly conflict. As National Security Adviser Zbigniew Brzezinski pointed out, the Soviet presence in Afghanistan was a chance to “give to the USSR its Vietnam war” (quoted in the work by Gibbs, 2006: 242; see also Gates, 1996: 143–149). Accordingly, the United States started funneling increasingly significant amounts of funding, weapons, and equipment to various Afghan insurgent groups. The goal was both to prevent a further expansion of Soviet influence and to bleed the Red Army white.

A more contemporary example of such dynamics is the war in Yemen, pitting the Houthi movement from the north against supporters of successive governments in the south. The root causes of the conflict are clearly domestic in origin. Over time, however, it has seen a steady increase in the involvement of outside powers, particularly by the two regional rivals Saudi Arabia and Iran (Clausen, 2015). In November 2009, Saudi Arabia launched a military incursion into northern Yemen to support government forces against the Houthis. By all accounts, Iran’s support for the Houthis was modest at that time. This, however, changed in the years that followed. Iran’s Islamic Revolutionary Guard Corps (IRGC) and its expeditionary Quds force began to develop closer ties with the Houthis, providing them with small arms, ammunition, and some financial assistance (Juneau, 2016; Terrill, 2014). Thus, in September 2014, when the Houthis seized the capital Sanaa, decision-makers in Riyadh were deeply concerned about the possibility of seeing a Tehran-friendly regime coming to power. A few months later, Saudi Arabia and a coalition of other Arab states initiated a military campaign involving massive air strikes and ground support for tribal militias opposed to the Houthis. The campaign’s main objective was to reinstall the country’s central government and, by doing so, to roll back Iran’s influence (Clausen, 2019; Juneau, 2024a).

In response, Iran ramped up its engagement in Yemen. Several United Nations (UN) reports have documented that in 2016–2017, Iran started providing anti-tank missiles, sea mines, radar systems, drones, and even parts for ballistic missiles to the Houthis (UN, 2017, 2018, 2019). Moreover, IRGC operators (augmented by Lebanese Hezbollah trainers) played a crucial role in helping the Houthis build local weapons factories and military training camps (Juneau, 2024b). Until further documentation is available, it is difficult to assess the exact motives behind Iran’s increased involvement in Yemen. The existing evidence strongly suggests, however, that one of the primary drivers was to bog down Saudi Arabia in an unwinnable war. Based on dozens of interviews with Iranian state officials, the usually well-informed International Crisis Group concluded: “Iran’s main objective in Yemen . . . became to ensure Riyadh got sucked into a deepening quagmire” (ICG, 2018: 23). ⁸

Abstracting from the details, the two cases nicely illustrate how a combination of perceived threats and opportunities can result in confrontational interaction patterns between rival powers in a third-party conflict. This type of proxy war aligns closely with Deutsch’s definition of the term, presented above, and receives considerable attention in current policy debates. Proxy wars, however, also feature other forms of rival power interactions.

Managed rivalry

A second type can be called managed rivalry. This means that rival powers support opposing sides in a third-party conflict, yet at the same time, they work together to reduce the risk of an inadvertent escalation. Of course, one could argue that all proxy wars are managed rivalries. After all, avoiding an outright military confrontation is typically one of the motivations for rival powers to act with and through local actors (Groh, 2019; Wiger and Atwell, 2023: 22–23). ⁹ Still, there is always the risk that proxy wars can escalate either vertically (to higher levels of violence) or horizontally (involving other states). In some cases, the desire to avoid such escalation dynamics is so great that rival powers take measures to actively de-conflict their operations in the local theater. Under such circumstances, the two rivals seek to strike a balance between, on one hand, backing their proxy partner, and, on the other hand, preventing the conflict from spiraling out of control. If push comes to shove, the latter objective (avoiding an escalation) trumps the former (improving the proxy’s battlefield situation). In other words, these are mixed-motive situations including incentives for both competition and cooperation.

Importantly, managed rivalries occur only when both external powers (or, to be more precise, the decision-making elites acting on their behalf) share escalation concerns. If this is the case, they are likely to take steps to manage their in-theater activities. This can include careful delineation of areas of involvement in conflict zones, establishment of communication channels, and advance notification of military operations. Moreover, external powers can commit themselves to limit the amount and type of resources (military supplies, logistical assistance, intelligence, combat troops) that they provide to their respective proxy partner (Anderson, 2019: 695–696; Lauren, 1983).

An example from the Cold War is the Angolan civil war (1975–1991). The Soviet Union and Cuba supported the People’s Movement for the Liberation of Angola (MPLA), while the United States and South Africa backed the National Union for the Total Independence of Angola (UNITA). In a masterful study, Anderson has shown that both sides established a range of practical de-confliction mechanisms. Most importantly, while the United States and the Soviet Union deployed military advisers, they shied away from deploying combat troops to Angola. Instead, they relied on South Africa and Cuba, respectively, to provide boots on the ground. Moreover, the intervening powers created geographic buffer zones and agreed to only target local forces—but not each other. All these measures were taken with the deliberate intent to avoid direct confrontations that could spark a wider regional war (Anderson, 2019: 701–704; for background on the war, see Groh, 2019: 157–181; Mitton, 2023: 197–228; Mumford, 2013: 48–51).

Another, more recent example is the Syrian civil war, in which the United States backed the Syrian Democratic Forces (SDF) while Russia (along with Iran and Lebanon’s Hezbollah) threw its weight behind the Assad regime. Shortly after Russia deployed fighter jets to Syria in September 2015, the Russian Defense Ministry and the Pentagon began to work out de-confliction procedures to reduce the risk of air-to-air encounters. This led to the signing of a memorandum of understanding that comprised air safety protocols, the use of specific radio frequencies, and the establishment of a direct communication line (DOD, 2015; Weiss and Ng, 2019: 4).

By summer 2017, US-supported rebel groups and Russian-supported government forces began to come uncomfortably close to each other on the ground, too. Thus, in a series of meetings between high-ranking US and Russian officials, additional de-confliction mechanisms were established. This included a geographical dividing line for ground operations, which broadly followed the Euphrates River. The US-backed SDF restricted its operations to the area east of the river, while Russian-backed government forces would remain west of it. Moreover, the two sides agreed to establish 24-hour communication links, including a separate email channel and a special phone line, connecting US officers with their Russian counterparts. The United States even created a specifically designated Ground De-Confliction Cell at the Ali al-Salem base in Kuwait, consisting of a senior military officer and four interpreters (Hamilton, 2018). In short, escalation concerns evidently played an important role in how US and Russian officials managed their countries’ involvement in the Syrian civil war.

Competitive influence-seeking

A third type of rival power interactions is best described as competitive influence-seeking. This means that two rival powers pursue highly adversarial strategies in a local conflict theater. Just like in confrontational proxy wars, they view each other’s involvement in zero-sum terms; one power’s gain is the other power’s relative loss. However, the two rivals do not support opposing warring parties. Instead, they vie with each other for the allegiance of one and the same local actor. There are several reasons why this may be the case. For example, both have strong ideological, ethno-political, or historical bonds with the same actor. Alternatively, both rival powers may have concluded that one warring party stands to prevail in the local conflict; hence, from a power-political perspective, it makes sense to establish close ties with the likely winner. Either way, the key point here is that two rival powers seek to outcompete each other to become the primary security partner and patron of the same local actor.

Drawing on the broader statecraft literature (for a useful overview, see Meierding and Sigman, 2021), we suggest that rival powers have at least three basic sets of tools with which they can attempt to draw a local actor into their orbit of influence. First, and most fundamentally, they can seek to affect the cost–benefit calculus of the local actor’s leadership through the provision of weapons, logistical support, troops, and other combat-related assistance. Second, rival powers can prop up certain individuals within the local actor’s leadership (through financial assistance, for instance) who favor an alignment with the external sponsor. Third, rival powers can conduct information operations and public diplomacy campaigns. The aim here is to shape attitudes and perceptions of the local actor’s leadership and its constituency. These tools, of course, are not mutually exclusive but can be combined.

An illustrative example from the Cold War is the Nigerian civil war, which erupted in May 1967 as the eastern province of Biafra announced its independence. Seeking to crush the separatist insurgency, Nigeria’s government appealed to its traditional Western allies, above all Great Britain, to supply it with advanced weaponry. The British government, however, at first declined to do so, largely due to strong public and parliamentary opposition (Smith, 2018; Wyss, 2023). The Soviet Union saw this as a prime opportunity to build closer ties with the Nigerian government, and if possible, to draw it into Moscow’s orbit. As one Soviet diplomat put it, “We will take advantage of mistakes the British make—it’s only human” (quoted in the work by Matusevich, 2003: 119; see also Jorre, 1972: 179–184). In the following months, the Soviets started supplying the Nigerian government with a variety of weapons, including assault rifles, howitzers, and jet fighters (Hoagland, 1970). In parallel, the Soviet Union began to expand its economic ties with Nigeria. Among other things, it extended a USD140 million loan, earmarked for the construction of an iron and steel mill (Legvold, 1970: 330; Matusevich, 2003: 119). On top of that, Moscow engaged in a concerted public diplomacy campaign, which involved the opening of new cultural centers alongside film screenings and the distribution of publications praising the two countries’ newfound friendship (Matusevich, 2018: 202–203).

Unnerved by the Soviet Union’s growing influence, the British government adjusted its policy and decided to go along with the Nigerian leadership’s request to provide advanced military equipment and training (Jorre, 1972: 299–302). As Maurice Foley, the Undersecretary of State, put it behind closed doors, any delay of providing military training is “doing us no good at all in Nigeria. . . at a time when the Russians are making every effort to get a foot in the door” (quoted in the work by Wyss, 2023: 503). ¹⁰ Similarly, the Secretary of State for Commonwealth Affairs, George Thomson, stressed in the British parliament: “If we were to cut off our supply of defense equipment [to the Nigerian government]. . . we would, I believe, lose our capacity to influence the Federal Government” (quoted in the work by Smith, 2018: 141). In consequence, Great Britain and the Soviet Union came to support the same side in the Biafra war—but they did not cooperate with each other. Quite the opposite, they sought to undercut each other’s influence.

Similar dynamics played out more recently in Mali. In 2013, France intervened on invitation of the Malian government after the country’s northern part had been taken over by jihadist insurgency groups and Tuareg rebels. France, the former colonial power, served in effect as the primary security provider (Boeke and Schuurman, 2015). In the late 2010s, however, Russia began to work increasingly pro-actively to enhance its influence in the wider Sahel area and sought to wean away the Malian government from France. ¹¹ Toward that end, RT (formerly Russia Today) and Sputnik launched a massive media campaign in Mali, promoting anti-French and pro-Russian narratives. Concomitantly, pro-Russian civil society organizations began to emerge in Mali, such as the Group of Malian Patriots, which sought to popularize a closer relationship with Moscow. In addition, Moscow offered the leadership in Bamako material assistance, including military and economic aid (ICG, 2023: 8–10; Le Roux, 2022). These efforts, however, proved largely fruitless.

The situation began to change only in August 2020 when Mali’s government was removed from power in a military coup. The coup was accompanied by large street rallies, with demonstrators waving Russian flags and carrying placards with anti-French slogans. The flags and placards were reportedly distributed by the Russian embassy (Gräns, 2023). ¹² Not to be outdone, France launched its own information operation in Mali, using fake accounts on social media channels to criticize Russia’s involvement (Velluet, 2020). Moreover, after returning from a visit to Paris, the Malian transitional president, Bah Ndaw, reshuffled his cabinet and removed several high-ranking military officers known for their pro-Russian leanings. This decision, most likely encouraged by France, prompted a second coup that took place in May 2021 (Haidara, 2021; Le Roux, 2022). The new authorities, led by Colonel Assimi Goita, immediately adopted a much more Moscow-friendly stance. In the months that followed, the new Malian leadership and Russia signed a number of political, economic, and defense-related treaties. In December 2021, Russia deployed military advisers and Wagner mercenaries to Mali, where they served as a counterinsurgency force and as a type of Praetorian Guard for Goita and his entourage. In return, the Goita regime severed its ties with France and other Western states, which, as a result, withdrew their forces from Mali (Gräns, 2023; ICG, 2023; Jourdain and Tuma, 2022).

The details of this case (just as of the Biafra war) are intricate, but the broader point is clear: France and Russia were locked in a struggle for the allegiance of Mali’s central government. This form of competitive influence-seeking entails a pattern of behavior that is markedly different from proxy wars characterized by outright confrontations or managed rivalries, as described in the previous pages.

Tacit alignment

A fourth type of rival power interactions can be called tacit alignments. This means that two rival powers back the same faction in a civil war and, at least to some degree, coordinate their behavior. To be clear, tacit alignments do not imply an underlying harmony of interests. The overall relationship between the two rivals remains marred by mistrust and hostility. Nevertheless, under certain circumstances, even staunch adversaries have strong incentives to coordinate their behavior, implicit or explicit, and to support the same side in another country’s civil war. This is likely to be the case if and when two rivals face a shared threat that is considered severe enough to warrant a temporary alignment. A shared threat can be another major power, which engages in the local conflict theater. Alternatively, it can be a local actor, which poses an imminent danger to the security or other core interests (however defined) of the two rivals. Simply put, competing powers confronting a common foe may consider burying their hatchet in the local theater and backing the same proxy partner.

Such backing may include the provision of complementary material support (such as economic aid from one rival power and arms transfers from the other), practical military cooperation at the tactical or operational levels, and some form of collaboration in politico-diplomatic forums. That said, it is often difficult for rival powers to coordinate their activities openly. Rival powers risk incurring audience costs if such behavior is perceived domestically as unacceptable appeasement. They also face reputational risks abroad, as overt cooperation may undermine their credibility with allies in other regions. Cooperation is, therefore, likely to proceed through deniable back channels to avoid accusations of duplicity (Henningsen, 2024: 20–21). Moreover, as noted above, there remain feelings of mistrust and resentment between the two. Hence, if the shared threat has been contained or defeated, their alignment is likely to fall apart. In other words, tacit alignments do not rest on a firm foundation upon which to build long-standing security cooperation; instead, they tend to be short-lived and geographically circumscribed.

It is worth pausing here for a moment to address a potential objection. A skeptical reader might argue that categorizing tacit alignments as a form of “proxy war” stretches the concept too far. After all, rival powers not only cooperate with each other, at least temporarily, but also support the same side in a third-party conflict. This surely cannot be described as a proxy war, or so one might say. ¹³ Such a critique, however, is misplaced. Recall that proxy wars, according to the latest wave of research, are defined as situations in which one or several outside powers (benefactors) intentionally support one or several conflict parties (proxies) in another country against one or several local opponents (targets). Going by this conceptual logic, tacit alignments squarely fall within the definitional scope of a proxy war.

An illustrative example comes from the Lebanese civil war (1975–1990). In the spring of 1976, Israel and the United States worried about the prospect of the Maronite Christian militias losing the civil war to the Lebanese National Movement (LNM), an alliance of pan-Arabist groups that supported the Palestine Liberation Organization (PLO). Syria, the Soviet Union’s most important ally in the Middle East, championed itself as a stout defender of the Palestinian cause (Hinnebusch, 2003: 197). Yet the leadership in Damascus also feared that a PLO victory would lead to an Israeli invasion of Lebanon and thus undermine Syria’s influence there. Israel, meanwhile, was deeply concerned that the PLO would be able to establish a military presence on its vulnerable northern border. In other words, for different reasons, both Syria and Israel wanted to contain the PLO’s influence in Lebanon (Sorby, 2011: 205).

Thus, by January 1976, Syrian and Israeli national security officials started exploring the possibility of a more coordinated involvement in Lebanon’s civil war (Levey, 2022: 775). Most negotiations were conducted through backdoor channels via the United States to keep them out of the wary eye of the public in the two countries. Unsurprisingly, there was also deep mistrust between the leaderships in Tel Aviv and Damascus. In the end, Israel accepted a Syrian military intervention on the condition that Syria abstained from operating in southern Lebanon and kept the military operations limited. Syria, for its part, agreed to the continued supply of weapons from Israel to the Christian militias, who in turn invited Syria to intervene (Levey, 2022: 782–783). Thus, in June 1976, Syrian forces intervened in Lebanon against the LNM and the PLO, thereby giving the Christian militias the upper hand in the civil war (Rabil, 2003: 53; for some background, see Mitton, 2023: 159–196). ¹⁴

Another, more recent case is the tacit collaboration between Iran and the United States in Iraq. In the summer of 2014, the rapid advances of the Islamic State (IS) caused a shudder to run through the spines of decision-makers in Washington and Tehran. Both saw IS and its self-declared caliphate as a major threat that needed to be countered (Laipson and Ollivant, 2024: 102–103). Hence, Iran and the United States decided to pause their rivalry for influence in Iraq and made common cause, despite their long-standing enmity. To be sure, Ayatollah Khamenei, the Supreme Leader of Iran, publicly precluded any formal cooperation with Inherent Resolve, the US-led military operation to defeat IS (Malakoutikhah, 2020). In response, Secretary of State John Kerry walked a fine line when declaring that against IS “There is a role for nearly every country in the world to play, including Iran,” but at the same time ruling out that Iran could be part of the US-led counter-IS coalition (BBC, 2014). Notwithstanding such pronouncements (which were likely aimed at placating hardliners in both countries), Iran and the United States started to work in tandem to support Iraqi government forces.

One example is the use of unencrypted radio communication on frequencies that both sides knew the other listened in on. In effect, US and Iranian-backed forces were well informed about each other’s movements. The use of Iraqi command centers as an intermediary is another example, which allowed both sides to convey important operational information (Cooper, 2015). At times, there even appear to have been concerted battlefield operations. In August 2014, for instance, in an effort to break a siege by IS of the northern Iraqi town of Amirli, Iranian-backed Shia militias along with Iraqi government forces took the lead on the ground. The United States, meanwhile, provided air cover (Cooper, 2015, see also Mansour and Jabar, 2017: 21). This was hardly a coincidence. As two seasoned observers pointed out, “Iranian ground operatives and Shia militiamen transmit target information to the Iraqi security forces, which in turn, share it with US forces” (Tabatabai and Esfandiary, 2017: 140). In other words, both Iran and the United States supported the Iraqi government forces and, at least to some extent, coordinated their in-theater activities. Of particular note is that this form of coordination continued even after the coming to power of the Trump administration and its hardline policy vis-à-vis Tehran. The tacit US–Iranian alignment started to fall apart only in late 2017 when IS had lost all its territory in Iraq.

These two brief accounts obviously cannot do justice to the complexities of each case. They nicely illustrate, however, that the emergence of a common threat can induce rivals not only to back the same proxy actor but also to align their in-theater activities. What this suggests is that even as we are moving into an era marked by intensified great power competition, one is likely to find instances of cooperation among rival powers in local conflict theaters.

Conclusion and directions for future research

The study of proxy wars involving rival powers has produced many important insights, but it has remained to date relatively disjointed and piecemeal. Different types of proxy wars have been addressed in diverse literatures that often do not, or only marginally, overlap. In this article, we have sought to integrate previously fragmented findings and to put them into conceptual order by offering a typological framework. This framework rests on two central choices that rival powers face when engaging in a third-party conflict. The first is whether to back the same or opposing local actors; the second is whether to coordinate their in-theater activities. Combining these two dimensions yields four distinct patterns of interaction: (1) confrontation (rival powers support opposing warring parties and engage in a highly adversarial behavior in the local conflict theater); (2) managed rivalry (rival powers support opposing warring parties but engage in deliberate de-conflicting); (3) competitive influence-seeking (rival powers support the same warring party but seek to outcompete each other); and (4) tacit alignment (rival powers support the same warring party and coordinate their involvement in the local theater).

While there is always the risk of exaggerating the payoffs associated with new typologies, we contend that our framework is useful in at least two respects. First, it offers a heuristic tool for systematically mapping interaction patterns between rival powers in third-party conflicts. In doing so, our framework provides the basis for structured between-case comparisons and large-n analyses aimed at uncovering recurring conditions associated with distinct types. Second and related, it facilitates finer-grained analyses of the dynamic nature of proxy wars involving rival powers. Take, for instance, the interaction between Iran and the United States in Iraq, which went from confrontation to tacit alignment and back again. Such shifts over time have received little systematic attention in the existing literature, which tends to treat proxy wars as static or discreet events. ¹⁵ In short, our framework provides the conceptual vocabulary needed to capture variations across space and time.

Where might research go from here? A promising avenue for future work would be to explore more thoroughly the question of what determines varying interaction patterns among rivals. In this article, we highlighted the role of risk assessments and threat perceptions but stopped short of advancing a full-blown theoretical account. Future work could move beyond this initial step and develop more specific propositions about when and why varying interaction patterns are likely to emerge. For example, building on realist scholarship, one might investigate whether bipolar international or regional orders are likely to see a higher rate of confrontational proxy wars than multipolar ones. Building on constructivist scholarship and the foreign policy analysis tradition, scholars could likewise examine whether and how features of the local conflict theater (e.g. history, culture, number and type of warring parties) and unit-level characteristics within rival powers (e.g. regime type, ideology, political party constellations) affect their interaction dynamics. In doing so, one would not only gain a better understanding of the factors shaping the strategic relationship between rival powers in local conflicts but also connect the study of proxy wars to broader theoretical debates in the field of international relations.

Another potentially promising research direction would be to study how different interaction patterns among rivals affect the duration and severity of local conflicts. The broader civil war literature suggests that one-sided interventions—where only one local conflict party receives outside support—tend to shorten civil wars, while two-sided interventions—where both local conflict parties receive outside support—tend to prolong them (see Balch-Lindsay and Enterline, 2000; Norrevik and Sarwari, 2021; Sawyer et al., 2017). Building on these findings, we would expect that rival powers pursuing competitive proxy warfare strategies increase the average duration of civil conflicts. After all, as described above, rival powers engage in an action-reaction pattern providing their proxy partners with material aid over an extended period. This, in turn, strengthens the capability and resolve of the local conflict parties to continue fighting. Similarly, we would expect that managed rivalries are associated with longer average durations of civil wars, though for different reasons. Here, rival powers provide their respective proxy partner with military support while they simultaneously agree to de-conflict their in-theater activities. This is likely to perpetuate the local conflict, as neither power is willing to provide their respective proxy partner with the support needed for a decisive battlefield victory. The likely result is a stalemated conflict that drags on (for similar arguments, see Anderson, 2019; Regan, 2002).

The implications are different, however, when it comes to competitive influence-seeking. As rival powers vie for the allegiance of the same local actor, they are eager to outcompete one another by offering significant amounts of material and military aid. In effect, the military balance of power on the ground is likely to shift decisively, enabling the favored actor to prevail on the battlefield and/or to coerce local opponents to make wide-ranging concessions. Other things being equal, this should decrease the duration of a civil war. Tacit alignments are likely to have a similar effect; when rivals jointly support and coordinate on behalf of the same local actor, that actor is better positioned to prevail quickly or impose a settlement.

Admittedly, these are broad-brush propositions that need further refinement and empirical testing. Still, they have a certain a priori plausibility and, as such, are suggestive of the typology’s analytical purchase beyond its immediate remit. More generally, we hope that our typology can serve as a stepping stone and source of inspiration for more sustained research on the linkages between great power competition, proxy conflicts, and civil wars. Much remains to be done.