Introduction

The idea of using proxies against adversaries has been familiar to strategic thinkers and practitioners for centuries. The practical application of both the “divide and rule” principle and the “my enemy’s enemy is my friend” principle was developed especially in the nuclear era, becoming an instrument of superpower competition as much as a means of avoiding direct military confrontation. During the Cold War, regional interstate armed conflicts and civil wars in Asia, the Middle East, Africa and Latin America were inspired, fuelled and controlled by superpowers pursuing their own strategic interests. The end of bipolar rivalry had temporarily reduced proxy wars to the margins of international security, but the 21st century has witnessed a return of this phenomenon, not to mention a tendency for reduced interstate conflicts, an increase in the internationalisation of internal struggles, a multi-polar world order, the emergent role of regional players, an erosion of statehood, the privatisation of violence and the militarisation of non-state actors.

Modern warfare involves the use of so-called “grey zone” tactics to a greater extent than ever before. This includes “information operations, political and economic coercion, cyber and space operations, proxy support and provocation by state-controlled proxy forces” (Hicks et al., 2019, p. 7)¹. The course and outcome of contemporary armed conflicts are not only influenced by entities directly involved in armed clashes, but also by external actors providing resources to non-state players involved in conducting military operations. By supplying arms, training and financing local forces, deploying military advisers, installing private contractors, delivering foreign fighters or mercenaries, as well as providing intelligence and logistics assistance, external patrons try to ensure their client’s victory, thus achieving their own long-term, strategic goals. Indirect interference in conflicts enables governments to wage war without any formal declaration, whilst benefitting from a significant reduction in risks and costs, simultaneously maintaining an international position of “plausible deniability” (Voß, 2016, pp. 40–41).

Whilst aggression below the threshold of war is perceived as one of the key challenges to international security, it is equally a strategic instrument. “Grey zone” threats include covert and clandestine activities serving a wide range of strategic goals, e.g. regime change, annexation of territory, destabilisation of target opponents. Empirical analysis of proxy war cases indicates that the grey zone model has an equal chance of guaranteeing “strategic victory” as becoming a “double-edged sword”. One of the key challenges for external powers waging war by proxy is to avoid unintended, negative consequences (so-called “blowback”).

The primary objective of this paper is to address the question of how state patrons control non-state proxies fighting on their behalf. In order to do so, a number of areas require specific consideration: 1) the criteria determining the dynamics of a patron-proxy relationship; 2) factors influencing the selection process of proxy forces; 3) elements that increase or weaken control over surrogate forces; and 4) maximising patron benefits whilst reducing both risks and costs by using proxies.

The author applied a neo-realistic research paradigm in his work. Patron-proxy relations were presented through the prism of the Principal-Agent theory, and through a detailed analysis of selected case studies (Iran & Hezbollah, Pakistan & Kashmiri militias, the US & Syrian rebels, Russia & Russian separatists in Donbas). The author indicates that elements such as ideology, ethnicity or religion do not necessarily provide control over proxies, whilst the range of common goals and interests, and the level of the proxy’s dependency are crucial.

Although the phenomenon of proxy wars has been present in international relations for years, it is impossible not to notice that it is treated marginally by researchers of political science, security studies and strategic studies. In his work, the author referred to both classic texts covering the subject in question (including those written by Karl W. Deutsch and Yaacov Bar-Siman-Tov) and contemporary research into this phenomenon (including that conducted by Chris Loveman, Geraint Hughes, Idean Salehyan, Andrew Mumford, Milos Popovic, Vladimir Rauta, and Armin Krishnan). Publications which did not cover the issue of surrogate warfare in a comprehensive manner, but were limited to single case studies, had an important supplementary value. In the analytical part of the work, press materials were also used, as well as analyses and reports from think tanks and research centres.

The strategy of war by proxy

The strategy of war by proxy is the art of influencing the course and outcome of conflicts in accordance with the interests of the third party by supporting proxy force(s) without the need for direct military intervention; all the more pertinent in the nuclear era where the cost of “total war” is beyond imagination. A clear reduction in risk permits proxy wars to be defined as being conducted “on the cheap”. Whereas they might not always be financially “cheap”, the strategic cost is usually lower than being actively involved in a nuclear conflict or a full-scale conventional war.

In the middle of the Cold War, Karl W. Deutsch understood the proxy war as “an international conflict between two foreign powers, fought out on the soil of a third country; disguised as a conflict over an internal issue of that country; and using some of that country’s manpower, resources and territory as a means for achieving preponderantly foreign goals and foreign strategies” (Deutsch, 1964, p. 102). A similar approach was presented by Yaacov Bar-Siman-Tov, who sees proxy conflict as a “war between regional states that may be regarded as a substitute for direct confrontation between the super-powers” (Bar-Siman-Tov, 1984, p. 263). These extremely general definitions were clearly determined in the context of the Cold War, when, as part of a bipolar balance of power, the United States and the Soviet Union were avoiding a direct confrontation by moving it to the territory of the “Third World”. Proxy wars between geostrategic powers took place in geopolitical “shatterbelts” in the Middle East (e.g. Lebanon, Israel-Palestine, Afghanistan), Southeast Asia (e.g. Korea, Vietnam, Cambodia, Laos, and Burma), Africa (e.g. DRC, Angola, Somalia, Ethiopia, and Eritrea) and Latin America (e.g. Guatemala, Cuba, Nicaragua, Salvador, and Colombia), where both sides sought to achieve geopolitical advantage by expanding their sphere of influence, or limiting such possibility to the opponent (Cohen, 1963).

As Andrew Mumford noted that “since the end of the Cold War, superpower-induced proxy wars have largely been replaced by proxy wars driven by regional powers via the cross-border percolation of militia groups, witnessed especially in Africa. The result is a shift in the character of these wars from internationalised conflicts of an ideological nature to regionalised interventions motivated by inter- and intra-state competition for power and resources” (Mumford, 2013a, p. 45). In the years between 1950 and 2010, as many as 89% of African states conducted this type of regional warfare, with primary sponsors including Angola, Chad, Ethiopia, Eritrea, Liberia, Libya, South Africa (in apartheid times), Sudan, Uganda, Rwanda, and Zaire/DRC (Craig, 2012, pp. 89–92). Numerous proxy wars involving regional players can also be witnessed throughout Asia (especially in the Middle East). After 9/11, Afghanistan (“the graveyard of empires”) once again became a proxy battlefield between global and regional powers just as it had been during the Cold War (Long, 2016). The situation in Kashmir is equally unstable, as Pakistan’s principal intelligence agency (ISI) continues to support separatist and Islamic terrorist groups in India (Behera, 2001). In 2004, following the 2003 US-led military intervention in Iraq, neighbouring Iran sponsored a proxy war against “The Great Evil” using local Shia militias (Thurber, 2014). By the end of 2010, massive anti-regime demonstrations, commonly referred to as “The Arab Spring”, led to internationalised internal conflicts lasting nearly a decade; Syria (Hughes, 2014b), Yemen (El-Ghamari, 2015), Libya (DeVore and Stähli, 2018).

With the end of the Cold War and an increase in the role of non-state actors, proxy wars gradually moved from state-centric conflicts to asymmetrical warfare, with government forces on one side and paramilitaries on the other. Since the 1990s, more than 90% of all armed conflicts in the world have taken an internal character but, at the same time, they are often internationalised due to the direct or indirect involvement of external actors (Gates et al., 2016, p. 2). The increasing role of non-state participants acting as proxies in armed conflicts is the result of deliberate action on the part of state-sponsors, as well as a certain erosion of state sovereignty (expressly the loss of monopoly on the use of force). The evolution of this phenomenon was caused by the collapse of the bipolar world, globalisation and the political destabilisation in certain geographical regions leading to the disintegration of state entities. The emergence of failed/fragile states, as well as the privatisation of wars (Kaldor, 2012), contributed further.

Changes taking place in both the international security environment and in methods of warfare have led to the evolution of proxy interventions, understood by Chris Loveman as an indirect engagement in a conflict by third parties wishing to influence its strategic outcome. In this approach, war by proxy is a kind of relationship between the external actor (benefactor) and the proxy (beneficiary), who can be either a state or non-state recipient of armaments, training, financial support, intelligence, logistical assistance and military advice (see Loveman, 2002, pp. 29–48). This type of indirect conflict intervention is aimed at maximising possible benefits whilst minimising risks and losses. Proxy wars allow states to pursue their strategic interests without the need to engage directly in politically, and financially costly, bloody warfare. The state directly initiating a conflict can be condemned by the international community and foreign sanctions may be imposed on it. In waging a war by proxy, politicians have no need to convince public opinion as to the merits of engaging their military in armed conflict. Neither do they need to bear responsibility for casualties nor incur significant financial resources related to the transfer of a large number of people and equipment into the war zone. Furthermore, their actions not risk political delegitimisation in the course of either a protracted conflict or during periods of armed struggle interspersed with ceasefire.

The structure and dynamics of the sponsor–proxy relationship

Vladimir Rauta indicates that “the structure of a proxy conflict involves three types of actors (state and non-state entities) whose roles are 1) benefactor, 2) proxy agent and 3) target agent. In this configuration, the relationship between the benefactor and the target agent is defined through an interest incompatibility over an issue and is conditioned by the impossibility of direct confrontation” (Rauta, 2013, pp. 254). As a result, this leads to a situation in which the “benefactor” conveys the burden of war on the third party (i.e. non-state “proxy agent”) to directly (militarily) engage against the “target agent”. As Rauta emphasises, “strategic interactions between a benefactor and their clients flow in two pathways: 1) from the state towards the non-state actor, and 2) from the non-state actor towards the state”. In summarising, we can distinguish three overlapping dyads between the participants of proxy wars: a) the beneficiary-target dyad, b) the beneficiary-proxy dyad, and c) and the proxy-target dyad (Rauta, 2018, pp. 456–468).

Given the accepted criteria relating to international viewpoints on the topic, Andrew Mumford distinguished four types of interactions between state and non-state proxy war participants: 1) a state uses another state as a proxy, e.g. the Israeli-Arab wars in the 1960s and 1970s waged with US and USSR patronage (Konyukhovskiy, Theocharis, 2019), 2) a state actor uses a non-state entity as a proxy force (e.g. rebels, resistance, insurgency, terrorist organisation, militia group or private military company [PMC]), 3) a non-state actor uses a state as a proxy; 4) a non-state entity uses another non-state actor as a surrogate (Mumford, 2013b, p. 45). Geraint Hughes (2014a, p. 12) limits the scope of sponsor-proxy relations to the “second variant” in which states assume the role of benefactor and non-state actors perform the function of surrogates, thereby reflecting trends in contemporary armed conflicts. Hughes further stresses that the patron-client relationship must meet three conditions: 1) the transfer of military support to the proxy; 2) a common goal (e.g. defeat the enemy); and 3) the relationship’s durability (at least for a few months). According to Mumford, limiting non-state actors to fulfil the role of proxy is “too static” and has an overly “state-centric” approach. He argues that the modern security environment enables non-state entities to pursue their own interests through other non-state actors and the states themselves. A similar position on this issue is taken by Assaf Moghadam and Michel Wyss, who noticed that “in recent years, an ideologically and geographically diverse set of non-state actors has adopted sponsorship roles akin to those traditionally held by states” (Moghadam and Wyss, 2020, p. 120).

In principle, the occurrence of one of the above models of relationship can occur when a non-state actor achieves a high level of autonomy (e.g. Hezbollah), but to some extent the proxy always relies on external support (Iran in this case) and its activities remain co-ordinated by the patron. It is difficult to agree with the arguments proposed by Mumford, Moghadam, and Wyss regarding the use of non-state proxies by “non-state sponsors”. Co-operation and partnership among terrorist organisations, armed groups and criminal organisations is a phenomenon well known to strategic studies (see Wiliams, 2008). The degree of asymmetry of power between them is incomparably lower than in the case of the state/non-state actor relationship. If, for example, we recognise the training, logistics, financial assistance and combat support offered by Palestinian organisations to leftist terrorists, and separatists, in the 1970s and 1980s as a sponsor-patron relationship (similarly the inter-group co-operation within the global jihad movement led by Al-Qaeda or ISIS and provided to their affiliates’ organisations), then the definition frames of “proxy war” would be distorted.

From proxy to sponsor? The case of the Hezbollah–Iran relationship

Some researchers believe that having gained such a level of autonomy and a strong position in their relationship with their patrons Iraq and Syria, the Lebanese Hezbollah could be perceived as an independent entity (El-Hokayem, 2007). The Party of God is certainly one of the most powerful and influential non-state actors in the Middle East; it is also a crucial Iranian proxy force in subversive operations against Israel, as well as shaping politics and security in Lebanon, Syria, Iraq, Yemen and Bahrain. Since the creation of this Lebanese paramilitary in the early 1980s, the group evolved from an “externally controlled subordinate” to a “strategic partner” for both Teheran and Damascus. As Kristina Kausch underlines, “for almost four decades of existence Hezbollah has undergone a significant transformation: a political transformation from a marginal political group into a party, a social transformation from a charity into a governance and social security apparatus, and a military transformation from a militia into a regional army and Lebanon’s most sophisticated military force” (Kausch, 2017, p. 40). The war initiated by this Shia organisation against Israel in 2006 can be taken as evidence of Hezbollah’s high level of political and military autonomy. Nevertheless, the group is not completely independent of external support. Although the Party of God has its own funding sources (including drugs and weapon trafficking, money laundering and donations), Iran remains the main sponsor of the organisation to which Hezbollah owes its present position (Caudillm, 2008, p. 129).

Even though Syria no longer enjoys the political and operational control over Hezbollah it had until 2000, this war-torn country plays a geostrategic role in the conduit of weapons, logistical support, and money. Teheran and Damascus have much broader political agendas than their proxy (who only partly shares common interests with its patrons). The convergence of interests combined with Hezbollah’s high level of organisation and structure, provide their sponsors (especially Iran) with few obstacles in co-ordinating military engagements. Since the mid-1980s, Hezbollah fighters have been a useful tool in forcefully representing Teheran’s strategic interests in the region; firstly by supporting “the export of the Islamic Revolution” and secondly, by influencing the regional armed conflicts in the Middle East since the first full year of the Arab Spring in 2011. The Lebanese terrorist group is, amongst others, responsible for the support of local proxies (mostly Shia militias) who undertake activities on behalf of Iran, rather than the Party of God’s leadership conducting activities themselves. We should therefore not define it as a “non-state sponsor”. Secretary General Hassan Nasrallah is undoubtedly able to use his organisation’s strengths to influence Iran’s Supreme Leader Ayatollah Khamenei and guarantee that Hezbollah’s strategic agenda will be taken into consideration. Nevertheless, due to the ongoing asymmetry of power, Hezbollah needs to consider Teheran’s stance and reaction in case of any autonomous activities.

The Principal-Agent framework and the problem of control

Relationship dynamics in proxy wars may be explained in the context of a “Principal-Agent” framework. This theory is based on the assumption that The Principal delegates particular tasks to The Agent in order to maximise performance. The model of war by proxy is based on the concept that a “sponsor state” will delegate war-related tasks to a non-state proxy (e.g. an insurgency or terrorist group, a militia, or private military contractor) in order to either achieve “plausible deniability” or to reduce costs relating to the direct use of force. This action in itself underlines the importance of the external actor’s role in shaping the conflict dynamics and exerting a level of control over the surrogate. Any external assistance (no matter how insignificant) to non-state actors involved in hostilities cannot be made public in case it is labelled as intervention by proxy. Milos Popovic, adopting the Principal-Agent framework in his research on proxy wars, explains that “in conflict studies, this usually translates into a government providing money, sanctuary, weapons or other tangible resources to rebel groups in return for their co-operation over goals, organisation and tactics”. In this manner, “the principal can select, monitor, and punish its agent by manipulating the provision of resources” (Popovic, 2018, p. 754).

According to the Principal-Agent theory, the delegation of tasks is a cost-saving mechanism and can be a useful tool if the benefactor lacks specific capabilities or expertise. But, as Idean Salehyan noted, at least two agency challenges may occurs: 1) adverse selection (or “information asymmetry”) – when principals do not have adequate information about the competence or reliability of agents prior to the establishment of a contractual relationship, and 2) agency slack (or “moral hazard”) – when the agent takes actions that are not consistent with the preferences of the principal (Salehyan, 2010, p. 495). However, as Salehyan underlines, “the patrons can minimise the risks by a) using tools (e.g. screening mechanisms) to select appropriate agents, b) monitoring the proxy’s actions through direct or indirect oversight mechanisms, and c) applying sanctions for disobedience” (Salehyan, 2010, p. 507).

Milos Popovic argues that non-state proxies often refuse to execute orders, desert from the battlefield, or turn against their patrons, because they are less centralised, formalised and accountable as their state patrons. Weak leadership (decentralised, factionalised) and a lack of internal control affects the agents’ decision-making process, resulting in proxies that are unable to swiftly adapt to changes in policy as a response to the principal’s demands. If the sponsor tries to impose its new course of war by proxy, non-state agents are likely to become internally divided in their approach to the external actor. If their loyalty and credibility would be questioned by the sponsor, the risk of a breakdown in co-operation increases. Despite the implementation of control mechanisms, principals often have to deal with problems in executing control over non-state proxies, because a) there are serious preferential differences between the principal and the agent, b) surrogates often do not reveal to their patrons all the information about their own objectives, capabilities, and actions, and c) sponsors abilities to efficiently monitor and punish proxy forces in case of transgression are limited. Popovic also notes that “there are three indicators of a proxy’s defiance, 1) the proxy may criticise a sponsor’s policy towards the armed conflict, 2) the proxy may refuse to carry out certain operations or twist the initial orders, and 3) the proxy may refuse to take part in negotiations or to sign a cease-fire, peace agreement, or other undertakings that are explicitly backed by their sponsor” (Popovic, 2017, pp. 922–925).

Overstretching the supply chain is a key factor in weakening or losing control over the proxy. The chain of supply may be short (e.g. the sponsor state’s intelligence and proxy leadership), or long (e.g. involving other actors and groups). The longer the chain, the higher the cost of monitoring and supplying resources (Popovic, 2017, p. 926). CIA agents discovered this when engaging Pakistani intelligence services (ISI) during Operation Cyclone, the US cover programme of military assistance to Mujahideen in Afghanistan in the 1980s. Due to the lack of its own intelligence assets, the CIA delegated the task of delivering weapons to regional partners who only supported factions they favoured. With only limited ability to oversee, monitor and control the entire operation, the Americans found that vast amounts of armaments were provided exclusively to radical pro-Pakistani warlords, or were sold on the “black market”. It is estimated that only 30-60% of US supplies actually went to the Afghan front (Behera, 2001, p. 397).

Criteria of proxy selection and control mechanisms

Relations between the benefactors and proxies are not a coincidence, but a combination of objective factors (i.e. the scope of common interests) and subjective factors (e.g. credibility, availability of alternatives, reputation, leaders’ intentions, precision of obligations, specific strengths, etc.). The involvement of external actors in proxy conflicts often involves the risk of making a mistake in selecting the surrogate, which may, in turn, lead to failure. As a consequence, the benefactor may fail to achieve their own strategic goals, becoming embroiled in an excessive pursuit of the proxy’s interests (“moral hazard”), or prompting long-term negative consequences (“blowback”).

Stephen M. Walt confirms that five rules determine the choice of an ally; 1) the actors form alliances against those who threaten them, 2) in order to minimise the threat from their adversary, they decide to form an alliance with him, 3) they choose allies with a similar ideology, 4) the element supporting the establishment of allied relations is the provision of foreign aid, and 5) the formation of an alliance is facilitated by the political penetration of a potential ally (Walt, 1994). Although the above regularities were formulated on the basis of interstate alliances, they are also reflected, to some extent, in the relations between state and non-state actors. However, the above elements vary in importance with some, e.g. common interests, playing a decisive role, whilst others (e.g. ideology, ethnic, religious or cultural ties) are additional. The durability and effectiveness of the sponsor-proxy alliance depends on the basis of its foundations.

The scope of common and divergent interests

The most important factors shaping the relationship between the patron and the proxy are the common goals they are striving to achieve (e.g. regime change, destabilising an adversary, expanding a sphere of influence, weakening or defeating a common enemy, balancing a threat from other actors etc.). Common interests naturally lead to cooperation, while discrepancies between allies (conflicts of interests) require compromise through negotiation; discrepancies alternatively being over-ridden by the stronger party as a result of the disproportionate levels of power between the patron and proxy. The Principal-Agent theory assumes that both “the principal and the agent are self-interested and that the agent may try to prioritise their own interests over the interests of the principal if given the opportunity” (Krishnan, 2019, pp. 544–545). Thus, the principal is primarily entrusted with the challenge of controlling proxies.

Relationships between participants of proxy wars are based on a calculation of losses and gains as an outcome of the co-operation. The higher the importance of the sponsor’s interests in a particular region, the higher the level of readiness to support the proxy. Such interests will include the low cost of providing assistance (both political and material), a high probability of success versus a low risk of failure, a low probability of negative consequences and the high worth of an ally in the pursuit of the sponsor’s interests (Salehyan, Gleditsch and Cunningham, 2011, p. 714). The latter element counting, among others:

The greater the potential of the proxy, the greater the prospect of victory. This does not, however, equate to an automatically greater likelihood of achieving the principals’ objectives. Proxies with a high potential are less dependent on external support and possess a stronger negotiating position (especially in the event of specific advantages), meaning they may be less willing to support the patron’s separate interests. The smaller the proxy potential, the greater the proxy’s readiness to accept external assistance on which they subsequently become dependent. In such instances, the proxy is more readily inclined to pursue the sponsor’s interests and hence more susceptible to the sponsor’s influence and control; the high level of a principal’s credibility, a lack of viable alternatives, a high threat level combined with an unstable, internal proxy environment all favour the sponsor. Milos Popovic notes that “if alternative support for the surrogate is available, the sponsor will have less leverage. When there are no alternatives, the patron will have greater leverage and the rebels will be more compliant” (Popovic, 2017, p. 928). We can conclude that according to the Principal-Agent theory, proxies with multiple sponsors have a greater leeway in their actions compared to those relying on a single patron.

Agents with “specific advantages” that the principal lacks, are able to create “information asymmetry” and use it to maximise their own interests, which may adversely affect the achievement of the primary goal (as shared with the patron) (Krishnan, 2019, p. 545). Therefore, the sponsor should not rely solely on the information provided by the agent, but constantly monitor their activities (e.g. by employing undercover liaison officers, advisors or intelligence operatives involved in the planning and implementation of the proxy’s operations) and apply a system of penalties and rewards (depending on the proxy’s effectiveness). Monitoring the proxy’s behaviour increases, to a degree, the principals’ costs, especially those associated with exposing the sponsor’s covert involvement; potentially leading to a lack of “plausible deniability” and increasing potential political consequences. But monitoring is crucial in maintaining a degree of control over the non-state proxy. A lack of monitoring may result in “strategic failure” and/or “blowback”.

In the case of violations or discrepancies in the declared and actual interests of the allies, the patron needs to have the possibility of sanctioning the surrogate for insubordination (e.g. by suspending the supply of weapons, financial support or depriving them of a safe haven). The ultimate tool for subjugating a proxy is the threat of abandonment; a threat that needs to be credible as the alternative for the patron might be a costly one. It should be noted that the effectiveness of all the above measures depends on the availability of alternatives and the level of self-sufficiency and the (relative) independence of the proxy (Salehyan, 2010, pp. 505–506). Withholding support or abandoning the proxy can lead to a breakdown of trust resulting in “blowback” (e.g. the defection of an agent to the patron’s adversaries, or turning against the former sponsor). The key element in managing the Principal-Agent relationship lies in the balance between providing proxies with assistance and allowing them to win battles, whilst exerting and maintaining political and operational control over them. Sponsors should invest in the development of proxy forces gradually, observing the extent to which the increase in combat capabilities translates into the implementation of the interests of the external actor, and to what extent the ally reacts and submits to the principals’ demands. The principal also needs to assess the proxy’s risk preferences and tolerance. The protection of a powerful external actor may encourage the agent to act more recklessly, e.g. the proxy may strive to conflict escalation beyond the limits acceptable for the patron in order to force him to increase the provision of resources needed to achieve a common goal (Krishnan, 2019, p. 546).

From the proxy’s perspective, the challenge is to maintain a balance between dependence on external support and (relative) organisational independence. On the one hand, it is in the surrogate’s interest to rapidly maximise its combat capabilities and overcome such challenges as the asymmetry of power, arms acquisition and securing sources of finance. On the other hand, in accepting foreign sponsorship it is important for the proxy to maintain a high level of autonomy in order to implement their own political agenda. As the Principal-Agent theory suggests, however, resources provided by the patron almost always reduces the autonomy of the armed group. Therefore, non-state proxies face a tradeoff between improving their resource base and maintaining organisational independence. This is crucial in the context of public support and the ability to mobilise the population (e.g. recruiting fighters). In deciding to co-operate with an external sponsor, local forces run the risk of being discredited as a “puppet” controlled by a foreign power. An example of this is the loss of Iranian public support for the People’s Mujahideen (MEK) co-operating with hostile Iraq in the 1980s (Salehyan, 2010, p. 507).

Inter-agency rivalry: a case of Russian proxies in Donbas

Although military collaboration between various rebel groups involved in multi-party civil wars could support the joint effort of defeating an incumbent government (or to force it to make concessions), such groups often compete rather than co-operate.

Instead of joining forces to achieve a common goal (e.g. regime change, territorial gains etc.), paramilitaries focus on pursuing their own particular interests such as the control of supply routes, crucial natural resources or civilian populations. Inter-rebel squabbling over the distribution of goods and benefits weakens the joint effort of toppling the strategic objective (Nygård and Weintraub, 2015, pp. 557–559). The survival of such rebel groups depends primarily on their military capabilities and the ability to inflict losses on the enemy’s forces, deterring them from attacks and intimidating civilians into co-operation. The military assistance from state sponsors (the “outside option”) theoretically increases the military potential of a certain paramilitary group, and may lead to armed victory over their adversaries if the opposing belligerents do not benefit from any external support or assistance. Moreover, by managing arms supplies, and other forms of military assistance, the benefactor strengthens their control over the proxy. But proxy wars are not as straightforward and this becomes evident when the Russian proxy war against Ukraine is analysed, specifically in comparing the annexation of Crimea with the separatist rebellion in Donbas.

Whereas in the case of Crimea, Russia conducted a direct covert action undertaken by Russian military units without distinction (special operations Spetsnaz and military intelligence GRU units), and with little support from local “self-defence” units serving as Kremlin’s proxies, the situation in Donbas was more complex and implications differed.

The Crimean operation made it possible to hold an illegal referendum on the “independence of the oblast (region)”, subsequently legitimising its illegal incorporation into the Russian Federation in 2014. Despite Russia breaking a number of international agreements, the Crimean crisis proceeded smoothly with a passive response from both the Ukrainian authorities and the international community. Spetsnaz soldiers were responsible for the seizure of key facilities (critical infrastructure), as well as their command and control (C2) at tactical level. Meanwhile, GRU officials organised and co-ordinated the activities of local “self-defence” units. The GRU exercised operational control over the unconventional phase of operations, co-operation with paramilitary groups, liaison with special forces units and local political leaders (Jonsson and Seely, 2015, pp. 9–10, 20).

In contrast to the “smooth” annexation of Crimea, the Donbas scenario triggered an insurgency and the permanent destabilisation of eastern Ukraine. Fuelling and controlling the course of the conflict in Donbas is Moscow’s tool for exerting influence over the government in Kiev. The Donetsk People’s Republic (DPR) and Luhansk People’s Republic (LPR) are fully dependent on Russian support, but the Kremlin does appear to have some problems in effective control over local warlords.

Whereas Russian President Vladimir Putin admitted, on the anniversary of the capture of Crimea, that operations on the peninsula were carried out by Russian soldiers, Moscow has consistently denied the presence of its armed forces in eastern Ukraine. Moscow admits that Russian citizens are fighting in the Donbas, but that they are “volunteers”. In fact, the process of recruiting them is controlled by Moscow through military commissions, PMCs, veteran groups and Cossack organisations. Prior to “volunteers” being transferred to the war zone, they undergo military training in Rostov-on-Don, where they receive equipment, armaments (devoid of serial numbers) and uniforms (devoid of labels and manufacturers’ tags in order to prevent identification). This allows us to conclude that, in fact, the “volunteers” are mercenaries financed by Russia. Foreign fighters are also directed to the fighting in Donbass, including Serbs (apx. 300), Cossacks and Chechens (Vostok battalion). Moscow has consistently maintained its position that it is not involved in inspiring and powering the rebellion in eastern Ukraine, and Russian diplomacy emphasises that the conflict is internal, not international.

The activities of Kremlin proxies in the Donbas are co-ordinated and supported by Spetsnaz and the GRU, as well as regular units of the Russian army (Sutyagin, 2015). In order to conceal the presence of the Russian military, soldiers have been absorbed into separatist units. As a result, OSCE observers might conclude that it is indeed individual “volunteers” from Russian armed forces who are fighting in Ukraine. The control over the separatists is exercised by former (or active) officers of Russian intelligence, special forces, regular armed forces and internal security services, some of whom also perform political and administrative functions. Exercising operational control over such a diverse environment seems to be a serious challenge and an area of competition for individual militias and special services (especially GRU intelligence units and FSB security services). For instance, Igor “Strelkov” Girkin and Alexander Borodai (both former FSB officers) were forced to return to Russia as a result of over-stretching their autonomy and competing with loyal Kremlin protégés. Several unruly warlords (including Arsen “Motorola” Pavlov, Mikhail Tolstych “Givi”, Alexander “Batman” Bednov and the so-called president of the DPR, Alexander Zakharchenko, have been killed during attacks. According to separatists, they were carried out by Ukrainian intelligence, but there are many indications that the liquidation of these people was carried out by rival separatist factions at the request of the Kremlin (Warsaw Institute, 2017).

The conflict in Donbas has assumed the guise of a ”business” rivalry between local war-lords and the intelligence services controlling them. The GRU supervises the military structures, while the FSB supervises the “civil” and “security” structures responsible for internal issues. While Russia has managed to achieve a relative stabilisation between individual groups of influence in the DPR, the LPR remains an area beset with power struggles. An example of such rivalry was the overthrow by a group of armed individuals, of the leader of the LPR, Igor Plotnisky, on 25^th Nov 2017. As a protégé of Viktor Surkov and the GRU, Plotnisky was deprived of his authority and replaced by Leonid Pasechnik (the former security minister of the LPR and a protégé of the FSB). The current division of influence in the Donbas runs along the borders of both self-proclaimed republics – the GRU controls the DPR, while the FSB controls the LPR (Piechal and Strzelecki, 2017).

Ideology, Ethnicity and Religion

Ideological, ethnic and religious ties are not synonymous with operational control of a non-state proxy, with said control defined as “the exercise of command authority over the irregulars: designating their operational objectives and providing authoritative direction for accomplishing those objectives through the tactical application of force” (Biberman, Genish, 2015). Non-state actors might approach ideology pragmatically, adjusting and aligning their profile and political agenda with those of the patron, thereby increasing the probability of assistance. In certain cases, proxies undertake specific ideological shifts in order to appeal to their particular sponsor (e.g. the leftist ideology adopted by many insurgent groups in the “Third World” in order to receive assistance from the USSR during the Cold War). As Popovic explains, “rebel ideology depends on the sponsor’s ideology, and (…) in turn, depends on access to foreign support” (Popovic, 2018, p. 758).

Ethnic, religious and linguistic alliances represent some of the “screening mechanisms” used in the proxy selection process. In addition to sharing a sponsor’s preferences, or at least being perceived as doing so, spiritual and cultural bonds increase the likelihood of a proxy’s acceptance for external assistance. Importantly, common languages, reduced barriers of communication, and shared cultural understandings ease the process of screening, monitoring and sanctioning agents, thus reducing “agency slack” (Salehyan, 2010, pp. 505, 509). For example, such rules apply to Russia which militarily supported the separatist endeavours of the Russian-speaking population in Transnistria (Moldova), Abkhazia and South Ossetia (Georgia), Crimea and Donbass (Ukraine).

By contrast, analysing proxy wars in the Middle East and explaining the structure of conflict between state and non-state alliances in Syria, Iraq, Yemen or Libya solely through the prism of religious and ethnic divisions is a fallacy (Kausch, 2017). Working on this premise alone, it is impossible to explain the conflicts of interest among the Sunni camp states (e.g. geopolitical rivalry between Turkey, Saudi Arabia, Qatar and Jordan supporting various rebel groups in the region), or the co-operation of Shi’ite (Persian) Iran with Sunni (Arab) Palestinian groups like Hamas and Palestinian Jihad.

Ethnic or ideological ties facilitate establishing co-operation and perform a very functional (narrative) role in justifying such actions by political leaders. They may also dissuade the proxy from defection, but cannot be seen as a guarantee of loyalty or a mechanism of sufficient control. Fotini Christia argues that “alliances in multiparty civil wars are not driven by the politics of ethnic kinship, but rather are consummated to secure minimum winning coalitions that guarantee participants the largest share of the spoils possible, while also making victory achievable” (Fotini, 2012). Using quantitative analysis, she examined two case studies, Afghanistan (1978-1998) and the Bosnian civil war (1992-1995), proving that the choice of alliances in civil wars is determined by relative power, self-interest and pragmatism instead of ethnicity (e.g. Hazara–Pashtun, Muslim–Croat alliances).

Out of control ideology: a case of Pakistani proxies in Kashmir

Since the 1990s, Pakistani ISI intelligence services have supported Islamic fundamentalists in Kashmir and India. In order to strengthen its control over extremists, Islamabad created the United Jihad Council (UJC) umbrella organisation, uniting 13 ISI-sponsored groups. However, this procedure turned out to be insufficient and did not prevent Islamabad from experiencing “blowback”. For example, in the mid-1990s, the Muslim Janbaaz Force and Ikhwanul Muslimeen units sponsored by Pakistan had abandoned their principal and started co-operation with their nemesis, i.e. Indian forces. A separate ISI proxy, Hisbul Mujahideen, declined to respect a cease-fire with India (2000) in clear opposition of Islamabad’s wishes. On 13^th Dec 2001, several terrorists from Jaish-e-Mohammad (JeM), a Kashmiri militant group sponsored by Pakistan, infiltrated the Indian Parliament and killed six police officers. It’s highly likely that President Pervez Musharraf was unaware of this deadly operation as, at the time, he was conducting negotiations with India intended to de-escalate conflict. In response to the attack on Parliament, the Indian army initiated a major military mobilisation (Operation Parakram) and two nuclear neighbours were bought to the brink of war. JeM’s action has delegitimised the Kashmir struggle for independence and provided India with the argument that there is no difference between Pakistan-sponsored jihadists and al-Qaeda (Popovic, 2017, pp. 922–925).

Another (un)controlled Pakistani proxy is Lashkar-e-Taiba (LeT), a terrorist organisation co-founded by the Islamic militant Hafis Saeed in 1987, with the aim of confronting the Soviet army in Afghanistan. LeT was then the militarised arm of Markaz-ud Dawa-walIrshad (MDI), Hafis Saeed’s Center for Preaching and Guidance through which the ISI was financing the Mujahideen. In the 1990’s, LeT moved its activities to the territory of Kashmir, and subsequently throughout India; all the while benefiting from the protection and support of the ISI. LeT’s political agenda is not limited to the fight for Kashmir’s independence. In fact, most of its fighters are not even from the region, principally being foreign volunteers motivated by Islamic fundamentalism. The group’s short-term goal is to “liberate” all Muslim-dominated regions in India (e.g. Gujarat, Hyderabad, Andhra Pradesh) and establish Sharia rule. The organisation’s priority is, therefore, religious struggle (jihad), which has naturally led to establishing co-operation with al-Qaeda and other Islamic groups operating in the region (e.g. Jaish-e-Mohammed, Hisbul Mujahideen).

LeT’s long-term goal is to create an Islamic caliphate throughout South and Central Asia (Clarke, 2010, pp. 2–14). Many years of ISI support have contributed to LeT gaining a level of operational capability that affords it a degree of autonomy. This has been evidenced, inter alia, by LeT’s terrorist activity following the events of 11^th Sep. 2001; contrary to the official policy of Islamabad. At that time, Pakistan became one of the main partners of the US in the war on terrorism, receiving, for this purpose (in the years 2002-2009), military support in the region of USD 12.4 billion. President Pervez Musharraf also attempted to normalise relations with India which had become disgruntled with elements of Pakistani (radical) intelligence and the country’s military community. As a result of terrorist attacks against India and military activity in Afghanistan, Pakistani authorities recognised LeT as a terrorist organisation in 2002. The group remains responsible for 60% of all terrorist activities in India (Counterterrorismproject, 2020).

The factors that led to the (at least partial) loss of Pakistani government control over LeT activities include, among other things, 1) factional divisions in the Pakistani secret services, 2) irrepressible religious motivation (Salafism and Wahhabism) intent on promoting the spread of jihad and the proclamation of a caliphate, and 3) achieving a high level of autonomy by diversifying the sources of its external aid (e.g. financial support from the Gulf States, from sympathisers in the UK and through drug trafficking profits) (Clarke, 2010, pp. 24–33).

The methods used by ISI to create, monitor and control its proxies in Kashmir seem to be unconventional, risky and inefficient. As Milos Popovic noticed, “Pakistan supervised the rise and fall of the Kashmir insurgency, pitting Hisbul Mujahideen against the Jammu and Kashmir Liberation Front (JKLF), and later Lashkar-e-Taiba against Hisbul, fearing that a dominant Kashmiri organisation could take on a life of its own and make a compromise with India” (Popovic, 2018, pp. 755). The case of Pakistani proxies in Kashmir proves that having many agents does not guarantee strategic victory except for wreaking havoc in neighbouring states.

Out of control ideology: the case of US proxies in Syria

Despite historical experiences, the US strategy of selecting proxies in Syria (initially to topple the regime, then to fight so called “Islamic State” [ISIS]) was based on the fighters’ “moderate” ideological profile. However, the handing over of military equipment to al-Nusra Front (affiliated to al-Qaeda) by US-trained Syrian rebels revealed the serious risks of “outsourcing war”. In this case, the recruitment of proxies based on their ideology had been more costly than expected (Biberman, Genish, 2015). Without efficient control of their operations, “moderate” rebels were unable to topple the regime, or to defeat ISIS. Washington therefore decided to collaborate with the Kurds, a traditional US ally in the Middle East.

The US intelligence community has been discussing replacing Syria’s authoritarian regime with a moderate Sunni government since the mid-1980s. According to the information published by WikiLeaks, the State Department was sponsoring Syrian opposition groups from 2006 – five years before the so-called “Arab Spring”. In August 2012, President Barack Obama authorised a covert action programme with the core objective of supporting the Free Syrian Army (FSA). US assistance included $25 million for “nonlethal” aid and $64 million for humanitarian relief provided to the Syrian opposition. Washington also allowed their Middle Eastern allies to operate a “rat line”, a supply chain from Libya through Turkey into Syria used to smuggle weapons to Syrian rebels. In late 2012, the CIA Director David Petraeus criticised Gulf State intelligence officials (Saudi Arabia and Qatar) for a lack of co-ordination with the CIA and Jordan in terms of supporting Syrian rebels, whereby the probability of achieving a common goal had been reduced (Krishnan, 2019, p. 547). In 2013, President Barack Obama once again approved a CIA covert action programme (codenamed Operation Timber Sycamore) to train and equip over 5,000 Syrian rebels with funding of US $850 million. But the American concept of a war by proxy in Syria was based on three (flawed) premises:

The US priority was to train “moderate rebels” not religious extremists, thereby minimising the probability of “blowback”. As it transpired, CIA officials operating in southern Turkey and northern Jordan faced serious difficulties in recruiting suitable candidates. From 7,000 recruits, only 1,700 were positively vetted, with the rest having had previous ties to radical Islamic groups. The cost of training each of the fighters was as much as $4 million and by the end of May 2015, the US had spent $41.8 million under the programme (Wright, 2015). “Division 30”, numbering 70 rebels, was the first unit to complete training and be sent into the field of conflict.

The ideological basis for selecting a proxy was neither limited to its combat readiness or competence, nor the sponsor’s ability to exert control over the proxy. In fact, in their clash with al-Nusra fighters, the entire unit was destroyed; some rebels having been killed, others kidnapped and some voluntarily joining the ranks of extremists. Thus, American weapons (including anti-tank weapons) fell into the hands of terrorists (Gartenstein-Ross, 2017). By October 2015, the US officially suspended the programme due to its ineffectiveness; a mere 150 or so rebels had been in the war zone (Shear, Cooper and Schmitt, 2015).

Although American proxies in Syria were well-paid², they were reluctant to fight against Assad’s regime or ISIS, preferring to engage in combat with other rebel groups. Monitoring and control mechanisms introduced by the CIA and the Department of Defence, such as a) the need for US approval to conduct operations, b) evidencing and reporting operational outcomes, and c) limiting finance to specific operations, were ineffective and inefficient, resulting in practical “adverse selection”. Apart from combat passivity, fighter desertion to al-Nusra or ISIS remained the main problem. The CIA and Pentagon frequently blacklisted groups that joined jihadists and held back their supply of weaponry. As a result, even loyal proxy groups suffered from limited provisions of military equipment (Krishnan, 2019, p. 551).

With the establishment of the so called “Caliphate” in Syria and Iraq (2014) and the onset of Russian military involvement (2015), Washington’s efforts became focused on fighting against ISIS instead of toppling Assad. The Obama administration organised and led a global anti-ISIS coalition consisting of 60 states that militarily contributed to Operation Inherent Resolve. International forces conducted air raids and missile attacks on ISIS positions, closely co-ordinating their actions with those of ground forces. A priority for the US administration was to limit its involvement to air and fire support, without committing its own “boots on the ground”. In the event, Washington committed an additional US $500 million to a Pentagon-led paramilitary operations programme where the objective was to train and arm Syrian anti-ISIS fighters at training camps in Turkey and Jordan. In February 2015, Ankara received US $29.5 to train 2,000 FSA fighters. If this phase of the programme was to be successful, it was planned to train a further 15,000 fighters over the following three years. After the first year, the programme was abandoned due to the failure to recruit enough rebels. The US Defence Department had spent a total of US $364 million of which US $325 million had been assigned to equipment, 4,000 small arms, a thousand vehicles, infrastructure, airlifts, and other costs, as well as US $30,000 on each of the 180 graduated recruits (Krishnan, 2019, pp. 547–548).

In March 2016, the Obama administration announced a new programme of supporting the Syrian opposition by training small groups of rebel trainers to independently organise larger units (train the trainers). Another $250 million was allocated to this end, bringing the total budget of the “train-and-equip” initiative to $1.1 billion (Cooney, 2016). In July 2017, however, a decision was made to axe the programme.

For 4 years, the CIA had trained and equipped a total of approx. 20,000 “moderate fighters” which, in the opinion of the new Donald Trump administration, did not quantify the high financial and political costs (Balanche, 2017). According to the Centre on Religion and Geopolitics, approx. 60 per cent of Syrian rebels were jihadists and many of them had committed war-crimes (Krishnan, 2019, p. 550). Inefficient procedures for vetting recruits combined with limited mechanisms for operational control of proxy forces (despite their declared “moderate” ideology) resulted in the vast majority of them joining radical Islamic groups with greater potential and thus a greater chance of success (this included both ISIS and al-Nusra Front). A further 1,600-2,000 fighters reached agreements with Hezbollah and joined the militia sponsored by Iran (Rasmussen, Ma’ayeh, 2018). In the opinion of American and Jordanian intelligence, military hardware (including TOW anti-tank missiles) supplied by the US and its regional allies largely fuelled the black market, thus carrying a high risk of future “blowback”. The high level of corruption within allied, regional intelligence services led to officers selling US weaponry for personal benefit (Mazzetti and Younes, 2016).

Armin Krishnan argues that the chances of achieving US foreign policy goals in Syria had been low from the onset:

The failure of Operation Timber Sycamore, aimed at removing Assad from power, led the US administration to change their strategy in Syria by prioritising the fight against ISIS. In late 2015, Washington sent 50 special operation forces (SOF) to Syria in order to commence collaboration with Kurdish fighters in Syria (Yekîneyên Parastina Gel, YPG militia, the main rebel group in a coalition labelled Syrian Democratic Forces, SDF). In April 2016, the US sent an additional 250 SOF to advise to Kurdish fighters and to conduct “kill or capture” operations against the leadership of so called “Islamic State”. In April 2018, there were over 2000 American soldiers in Syria at approx. 20 bases, mostly in Northern Syria controlled by the Kurds (Krishnan, 2019, p. 548).

Although Syrian and Iraqi Kurds (Peshmerga), backed by forces of the global anti-terrorist coalition, have played a key role in fighting ISIS on the ground, the negative consequence of this policy was the deterioration of relations between Washington and Ankara. Turkey considers Kurdish separatism as their main internal threat. America’s NATO ally claims that there are direct institutional links between Syrian Kurds and the Kurdistan Workers’ Party (Partîya Karkerén Kurdîstan, PKK) considered to be a terrorist group by Turkey, the US and the EU. Americans have a long tradition of co-operation (or operational involvement) with Iraqi Kurds. Co-operation with the YPG (labelled as an SDF in order to limit political “blowback”) was established in particular due to the failure of paramilitary operations with Syrian rebels, as well as the fear of losing influence to Russia – tactically collaborating with YPG and the Iraqi Peshmerga (Kausch, 2017, p. 43). This led to a deterioration in the US-Turkey relationship, due to the fact that Ankara perceives YPG and YPJ militias (affiliated to PKK) as a terrorist organisations.

Conclusion

The delegation of armed conflict to non-state proxies is an age-old, and frequently-used instrument employed by states wishing to influence a strategic outcome without costly and direct military intervention. The Principal-Agent theory demonstrates that such an approach contains many attractive benefits, but equally underlines the potential problems and risks associated with the selection and control of agents.

The decision to initiate collaboration with a non-state proxy is influenced by factors such as the proxy’s potential, “specific advantages”, and the stability of the internal environment in the region concerned. In order to maximise potential benefits whilst minimising risks and financial commitments, sponsors need to evaluate the extent of key mutual interests shared with the agent in the decision making process. Ethnic, religious or ideological ties help in establishing a relationship with the proxy, but are not necessarily key factors in determining the effectiveness and outcome of the association.

The durability of relations between proxy war participants is primarily determined by the scope of interests shared by both the Principal and the Agent. In order to exercise control over the proxy, the patron dominates and manipulates the proffered support, thereby increasing the chance of achieving their own objective. Conversely, proxy forces strive to increase their combat readiness and potential, and gain (relative) independence.

All parties involved in the arrangement can exert additional influence by establishing co-operation with other entities involved in the conflict; both directly and indirectly. By expanding the scope of the “political manoeuvre”, either protagonist hopes to increase the chance of victory, whilst simultaneously running the risk of losing the “protector” and potentially leading to failure. In playing such “games”, actors need to continually re-evaluate the arrangement and calculate which actions will bring the greatest rewards, thereby allowing them to implement their own political agenda.

Principals cannot overlook the risks resulting from “adverse selection” or “information asymmetry”. Effective control mechanisms (e.g. the presence of military advisers, liaison officers etc.) must be in place in order to monitor and ensure that the proxy is acting in line with the principals’ interests. Proxy forces operating without supervision will be prone to deviating from the agreement, or even abusing their position (“agency slack” or “moral hazard”). This is all the more likely if “specific advantages” are to be gained. Such scenarios may lead to unintended negative consequences for the sponsor (“blowback”) and “strategic failure”, which principals need avoid at all costs.

In conclusion, Principal-Agent relations in proxy wars are to be viewed as potentially unstable informal agreements. Whilst initial benefits can be palpably measurable, subsequent outcomes are overwhelmingly prone to fragmentation and rarely result in achieving the mutually desired objective.