## Introduction

The use of proxies in warfare is typically understood as a state sponsor's reliance on military surrogates that are outside the purview of the state's conventional armed or security forces, and that offer services to their benefactors in exchange for tangible material support.1 A long-standing feature in the history of armed conflict, the reliance on surrogates has become particularly endemic in the post–World War II era, with important implications for international security.2 Following the attacks of September 11, 2001, and the ensuing “global war on terror,” the use of proxies has sparked renewed attention among academics and policy analysts alike, who have examined its causes, nature, and consequences in local, regional, and international contexts.3

Studying trends and patterns of sponsor-proxy relationships matters because they have wide-ranging repercussions.6 Existing studies have shown that the provision of external support to belligerents in civil wars, insurgencies, and other forms of political violence internationalizes these armed conflicts, raises their lethality rate, and increases the likelihood of conflict relapse.7 Exploring the evolving role of nonstate actors in this context is of particular importance given their potential to disrupt and destabilize regional and international security, as was most recently visible in the case of the Islamic State.8 The presence of nonstate actors raises the overall number of belligerents in a conflict theater, with many of them using indirect modes of warfare such as terrorism that confront states with vast military, political, financial, and legal challenges. Finally, many contemporary armed nonstate actors—including those that sponsor proxies—are transnational. Disrupting their efforts requires complex and costly international coordination on diplomatic, legal, military, intelligence, and humanitarian matters.

This study suggests that nonstate sponsors employ proxies in ways, and for reasons, that differ from those of traditional state sponsors, and with distinct implications for international security. Conventional insights on proxy relationships and proxy warfare hold that state sponsors employ proxies in an effort to advance their strategic objectives in a cost-effective manner, while minimizing the risk of becoming embroiled in a major military conflict. States utilize proxies to achieve both political and strategic objectives, but they conceive of proxies as an indirect, predominantly military tool to achieve these goals.

To examine how nonstate sponsorship differs from state sponsorship, we conduct a comparative analysis of three cases of nonstate sponsorship: al-Qaida in the Arabian Peninsula's (AQAP's) engagement with Sunni Bedouin tribes in Yemen; the People's Protection Units’ (YPG's) sponsorship of proxy groups under the banner of the so-called Syrian Democratic Forces (SDF) in northeastern Syria; and Hezbollah's sponsorship of the Lebanese Resistance Brigades in Lebanon. These cases are selected based on variation in the ideology and capacity of the respective sponsors, and in the degree of political support they enjoy in theater.

Theoretically, we argue that state and nonstate actors use different strategies when engaging proxies given the combined effect of endogenous traits and exogenous constraints, which apply differently to the two sponsor types. Endogenous traits include organizational capacity and objectives, whereas exogenous constraints pertain to distinct pressures and limitations that restrict the respective sponsor's ability to maneuver.

Compared to nonstate sponsors, states face relatively few capacity problems. They typically employ proxies to attain their regional or global aims, rather than to ensure their survival. At the same time, state sponsors often face domestic challenges to unwanted involvement in foreign wars, while internationally they are constrained by prevailing norms against intervention in foreign conflicts.10 As a result, state sponsors view and utilize proxies as an instrument to advance their strategic objectives while reducing domestic audience costs and the risk of international penalties. For state sponsors, relying on proxies as an open political instrument would, in most cases, defeat their overall objectives and exacerbate the constraints they face—hence their tendency to utilize proxies as military surrogates, as such collaborations can be denied more plausibly than relations with highly visible political partners.

Nonstate sponsors, in contrast, are typically plagued by a capacity gap that hampers the attainment of organizational objectives of survival and growth. Often, their quest for self-preservation is further complicated by two key exogenous constraints: a governance deficit and a legitimacy deficit. Nonstate sponsors, we argue, employ proxies to address their organizational shortcomings while seeking to reverse the deficits they face. These strategic requirements and challenges are primarily political, not military. For this reason, nonstate sponsors prefer partnering with local proxies that possess comparative political advantages that they themselves lack. As we show in our case study analysis, our argument applies to nonstate actors regardless of their ideology, military and economic capacity, or level of popular support.

Our findings imply that relationships between most nonstate sponsors and their proxies tend to be more symmetric than those between state sponsors and their surrogates. Not only do most nonstate sponsors face limits in their military capacities, but many also experience ongoing challenges to their organizational survival. For low- and moderate-capacity nonstate sponsors, in particular, the reliance on proxies is based more on need than on interest. This suggests that proxies of nonstate sponsors possess greater leverage over their benefactors than proxies of state sponsors do over theirs. Taken together, therefore, we can expect such arrangements to be more transactional and pragmatic, and less enduring. Broadly speaking, the use of proxies presents risks to their nonstate sponsors—they are more susceptible than state sponsors to pressures and manipulation by their proxies and external actors intent on undermining these relationships.

Nonstate sponsors such as Hezbollah, which possesses unusually large capacity, are a notable exception. Such high-capacity actors have great leverage over their proxies, and forge relationships that are more asymmetrical, akin to those between most state sponsors and their militant clients. That said, few nonstate actors are as potent a power broker as Hezbollah or the Islamic State at its apex. Hence, the majority of sponsor-proxy relationships involving nonstate sponsors are likely to resemble transactional arrangements.

Our study has several limitations. First, it examines a small number of cases—all from the broader Middle East region—and will hence require additional empirical testing before more robust conclusions can be offered. Second, our study endeavors to identify macro trends, which requires us to make generalizations that are broader than would ideally be the case. This limitation applies in particular to our discussion of state sponsors, which are not the main focus in this study and whose behavior is described in aggregate. Third, our study does not settle many of the conceptual and definitional questions related to nonstate sponsorship and sponsor-proxy relations more broadly. Instead, our main hope is to make an initial contribution to understanding core aspects of nonstate sponsorship—an issue that merits greater analytical attention.

The remainder of this article is structured as follows. The first section defines sponsor-proxy relationships and situates our study within the broader literature. The second section introduces our main argument, lays out the theory informing it, and presents the observable implications and hypotheses that follow from that discussion. In the third section, we use three case studies of contemporary nonstate sponsors to test our hypotheses: al-Qaida in the Arabian Peninsula, the People's Protection Units, and Hezbollah. In the conclusion, we offer some thoughts on the broader implications of our findings.

Although it is widely acknowledged that the use of proxy forces has a wealth of historical antecedents, their systematic study is a rather recent phenomenon.11 Despite the increased attention that the use of proxies in conflict has drawn in recent years, however, scholars have yet to agree on a general, integrated theory of proxy sponsorship.12 One major impediment toward the emergence of such a theory has been a lack of terminological and conceptual clarity related to the use of proxies in conflict.13

The present study places its analytical attention on the relationship between sponsors and proxies, thus taking an “actor-centric” approach.14 This approach contrasts with the more common usage of concepts such as “proxy wars” or “proxy conflicts,” which shift the focus of analysis to specific conflict theaters and to the presumably dominant strategy used in these conflicts.15

We define “sponsor-proxy relationships” as informal collaborative arrangements between asymmetrically capable parties, in which one party (the sponsor) utilizes another party (the proxy) to reach its strategic goals in exchange for tangible assistance.16 The types of assistance provided by the sponsor and the exact services rendered by the proxy differ from case to case, but the sponsor's assistance typically includes a combination of the following: provision of weapons and equipment; financial assistance; training; intelligence; operational planning; the provision of a safe haven; political cover; or some combination thereof. The proxy's services usually comprise a combination of the following: fighting a common adversary; collecting intelligence; patrolling and holding rear areas; and/or exerting governance on behalf of the sponsor.17

This definition situates sponsor-proxy relationships within the broader category of collaborative arrangements between at least two parties that involve the use of force (or the threat thereof) to attain political objectives. At the same time, it highlights two distinct features. First, sponsor-proxy relationships are less formal than conventional interstate alliances, which typically include mutual security guarantees and written agreements stipulating “the contingencies in which military cooperation will occur.”18 Second, the sponsor's privileged status distinguishes sponsor-proxy relationships from cooperative relationships among militant groups more broadly. The latter are not inherently and necessarily asymmetric; do not necessarily imply a subordinate role distribution between the involved parties; and do not always prioritize one party's strategic objectives over those of the other.19

### STATE SPONSORSHIP OF PROXIES: CAUSES AND COSTS

The existing literature on the use of proxies in conflict may be divided into Cold War and post–Cold War scholarship. Whereas the former has mostly relied on descriptive single-case studies, the latter has become increasingly diverse in its approaches and methods. Both share the tendency to attribute the role of sponsors almost exclusively to states.20 Broadly speaking, the extant scholarship offers three main reasons why states adopt sponsorship roles. First, sponsors use proxies because of their perceived military value, such as a superior knowledge of the local terrain or population, or specific tactical and/or operational capabilities.21 Second, state sponsors assume that utilizing proxies is a cheaper option than direct military action. This calculation includes direct costs associated with the deployment of armed forces as well as domestic audience costs (e.g., political constraints on military action, casualty sensitivity, and war weariness), and international condemnation or sanctions.22 Third, the use of proxies may offer plausible deniability to state sponsors wishing to obfuscate their involvement.23 Examining the use of pro-government militias as proxies, for example, Ariel Ahram argues that colluding with nonstate actors allows states to distance themselves from flagrant violence committed against civilians.24

Thus, there is broad agreement in recent studies that sponsors choose their proxies because of a relative disadvantage vis-à-vis their adversaries in terms of their military capacity or because they want to reduce costs associated with direct military action, or both. For some state actors with weak conventional military or security force capabilities, utilizing proxies may be one of the few options to project power toward an external enemy or to confront domestic armed opposition groups in remote areas.25 Scholars also largely agree on the potential costs of sponsorship: proxies may pursue divergent goals; divert resources according to their own preferences; engage in uncooperative behavior; devote suboptimal effort; or even switch sides and/or turn against their benefactors.26

To be sure, political motivations factor into states’ decisions to sponsor proxies, and at times may be the predominant driver of surrogate sponsorship. During the Cold War, for example, the superpowers availed themselves of proxies partly as an exercise of mutual “covert signaling.”27 Additionally, backing proxies can in some instances help shore up domestic support for the state sponsor.28 In practice, it is often difficult to neatly separate political from military motives for proxy sponsorship. Our reading of the literature does not deny the role of politics in driving state sponsorship of proxies. In aggregate, however, such a review reveals that, more often than not, states tend to value proxies for their real or expected military contributions than for their perceived political utility—and indeed, their proxies typically perform a primarily offensive military function. As the next section shows, when nonstate actors employ proxies, they reach the opposite conclusion about their militant client's value and utilize them in different ways.

## The Rise of Nonstate Sponsors

In recent years, there have been signs that some nonstate actors may perform superordinate functions in sponsor-proxy relationships (i.e., as sponsors that employ other nonstate actors as proxies). This argument supports the notion that nonstate actors possess greater agency in international affairs than is commonly attributed to them.29

The existence of nonstate sponsorship has been acknowledged in a handful of studies. Andrew Mumford, for example, describes the “benefactors” in sponsor-proxy relationships as “a state or nonstate actor.”30 Christopher Phillips and Morten Valbj⊘rn have distinguished between state sponsorship and nonstate sponsorship in the case of the Syrian civil war,31 and a recent report published by New America suggests that “the new and emergent political economy of conflict has empowered proxies themselves to develop their own proxies,” making contemporary nonstate actors “both principals and agents.”32 The above referenced studies, however, do not provide comprehensive case studies of nonstate sponsorship. Nor do they explore the causes, nature, and consequences of nonstate sponsorship.

### WHY NONSTATE SPONSORS EMPLOY PROXIES

Why and how do nonstate actors sponsor proxies, and to what effect? Whereas for state sponsors, proxies are a military means in the sponsor's quest to attain political ends, nonstate actors are guided by a rationale that is less military-centric.33 We argue that nonstate sponsors employ proxies primarily as political instruments, with their military contributions a secondary consideration. The primary functions of proxies of nonstate sponsors are as political ancillaries that service political goals in an unmediated, direct fashion. For nonstate sponsors, these political goals often center around the consolidation of their own political power, mainly by enhancing their quest for legitimacy—a minimal degree of which is a requirement for assuming and retaining power in the post-conflict phase.34 To these nonstate sponsors, which are often effective in combat but lack popular support beyond the immediate communities they claim to represent, the proxy's military support is not immaterial but a side benefit—one whose importance is secondary to the proxy's value as a political, legitimacy-enhancing asset.

Theoretically, we argue that state and nonstate actors employ these different proxy utilization strategies because of two sets of factors. The first relates to the endogenous traits of the respective sponsor, namely the sponsor's organizational capacity and its overall objectives. The second relates to exogenous constraints, namely distinct pressures and limitations that restrict the respective sponsor's ability to maneuver. These dynamics are summarized in table 1 and described next.

Table 1.

Proxy Utilization by State and Nonstate Sponsors: Theoretical Underpinnings

Endogenous TraitsExogenous ConstraintsProxy Utilization Strategy
State Sponsors high capacity; efforts to maintain/enhance regional and/or global strategic positioning domestic audience costs; international norms proxies as military surrogates
Nonstate Sponsors low capacity; efforts to ensure organizational survival and growth legitimacy deficit; governance deficit proxies as political ancillaries
Endogenous TraitsExogenous ConstraintsProxy Utilization Strategy
State Sponsors high capacity; efforts to maintain/enhance regional and/or global strategic positioning domestic audience costs; international norms proxies as military surrogates
Nonstate Sponsors low capacity; efforts to ensure organizational survival and growth legitimacy deficit; governance deficit proxies as political ancillaries

ENDOGENOUS TRAITS. State sponsors typically enjoy greater political, military, and financial capacities when compared to their nonstate counterparts. In general, state actors also enjoy greater legitimacy when compared to armed nonstate actors. Consequently, enhancing their legitimacy is not a primary concern for them, and hence unlikely to be a key driver of states’ decision to sponsor proxies.35 Instead, by employing militant clients, state sponsors seek to advance their regional and global strategic goals.36 Broadly speaking, state sponsors use proxies in the hope that the latter will affect the military outcome of conflicts in the sponsor's favor while avoiding direct military action by its own armed forces.

Compared to state sponsors, nonstate actors will typically have fewer financial and military means at their disposal, which exacerbates their political weakness when compared to that of state sponsors.37 Although a handful of nonstate actors enjoy great political and military clout, most face internal challenges to the extent that they are unable to prioritize the advancement of broader regional and global strategic objectives, at least until such time as their own survival has been ensured. Hence, they pursue goals that are mostly local or national in scope. Many seek to take over the reins of the state or to achieve regional autonomy or independence.38 Some of them, such as the Islamic State, may have transnational goals, but even transnational nonstate actors face challenges consolidating power in just one state.39 By definition, armed nonstate actors have not achieved the status of a state actor, and their aspirations will typically meet stiff military resistance on the part of a state or, as in the case of the Islamic State, a coalition of states.40 This reality turns self-preservation into the nonstate sponsor's most immediate need. Even nonstate actors such as Hezbollah, which are more secure in their power and may not face immediate threats to their survival, seek to enhance their political power and growth. They aspire to become more legitimate and representative actors, and to that end seek to present themselves as entities that are able to govern effectively, in state-like fashion.41 Utilizing proxies is a means for them to make that case before domestic and external audiences.

EXOGENOUS CONSTRAINTS. States and nonstate actors also utilize proxies differently because of exogenous constraints. These constraints restrict the sponsors’ maneuverability in different ways. They affect the selection criteria by which different sponsors identify suitable proxies, and they influence the sponsors’ decision on whether to openly acknowledge their relations.

When states seek to wage war, they face both domestic and international constraints.42 Domestically, direct military action often has negative repercussions, such as casualty sensitivity and war weariness.43 Internationally, states face a different set of costs to foreign entanglement.44 International penalties can range from verbal condemnations and diplomatic isolation to sanctions and, in extreme cases, military conflagration with nonstate actors or even other states.45

Nonstate sponsors, in contrast, have different sets of constraints. Domestically, they typically encounter fewer protestations over casualty rates than state sponsors and are less likely than states to tolerate complaints of war weariness among their ranks.46 On the contrary armed nonstate actors often pride themselves on their ability to fight, and even sacrifice their lives for their cause, which can account for part of their appeal, including the attraction of foreign fighters in some cases.47 Because armed nonstate actors need to safeguard their reputations as fighters, they will be more hesitant to delegate combat operations to proxies. Employing proxies for offensive combat operations could even backfire, as potential supporters might perceive such delegation as a sign of weakness and reduced morale on the part of a nonstate sponsor that had previously branded itself as a highly motivated fighting force. For this reason, nonstate sponsors are less likely than state sponsors to pass on the main responsibility for military operations to their proxies. To the extent that proxies of nonstate actors fulfill military roles, they are more likely to support rearguard duties such as security and patrolling functions.

Similarly, armed nonstate actors are less likely to be intimidated by the threat of international penalties. On the contrary, they often provoke incumbent regimes into an overreaction, hoping to draw support from the local community or international backers.48 Alternatively, they may goad international powers as part of a provocation strategy designed to draw foreign forces into a conflict, expecting that foreign occupation will draw recruits to their cause.49

Nonstate sponsors often face a different set of constraints to their political efforts. The first is a governance deficit (i.e., a limited capacity or an ability to provide basic services to the population under their control).50 Areas under the control of nonstate actors often lack institutions of governance for a variety of reasons, including a lack of resources, a permanent state of conflict, or lack of popular support. Even where those institutions are present or where high-capacity actors do provide basic services, the provision may be geographically uneven, or may be entirely interrupted given ongoing conflict and an unsteady supply of resources. When lack of popular support is endemic, it can exacerbate a second exogenous constraint, namely the nonstate actor's intheater legitimacy deficit. Armed nonstate actors need local support—or at the very least, local acquiescence—to ensure their survival, but often face a local population that is mistrustful of their intentions.51 To be sure, nonstate actors may be seen as legitimate by the communities they purport to represent. Nevertheless, many armed nonstate actors tend to pursue specific causes, and hence often appeal only to certain population segments. They frequently face inherent limitations and commitment problems in their effort to draw support from the broader population, because they represent narrow ethnic, sectarian, or ideological causes.52

Where and when one or both of these deficits exist, nonstate sponsors will identify and work through proxies as a preferred solution to address these shortcomings. To optimize their chances at success—and foremost their survival as a group—nonstate sponsors wish to come across as attentive to local needs and attuned to local norms.53 In countries dominated by certain ethnic, sectarian, or tribal traditions, sponsors will therefore tend to rely on proxies that represent these local identities. In countries characterized by ethnic heterogeneity, nonstate sponsors are more likely to select proxies that reflect such heterogeneity. Nonstate sponsors, in short, worry much more about how local communities perceive them. They choose proxies that can embellish their image and enhance their legitimacy. In contrast, state sponsors are, by and large, less concerned about such reputational costs.54

### IMPLICATIONS AND HYPOTHESES

Our argument has a number of observable implications that allow us to formulate testable hypotheses. We group these implications and the corresponding hypotheses into two categories, both of which have key relevance for our research questions: proxy selection and proxy utilization.

PROXY SELECTION. In contrast to state sponsors, which are likely to select their proxies based on their expectation of the latter's military performance, our theoretical discussion suggests that nonstate actors are likely to select their proxies according to their expected political utility. Therefore, we expect sponsors to select groups as proxies whose identities closely resemble those of the communities that the sponsor seeks to influence. Moreover, because the nonstate sponsors seek to maintain a reputation of military strength and high commitment of its combatants, we expect them to utilize proxies that are militarily weaker than themselves. Doing so has the added benefit of reducing the potential military threat that these groups might pose to a sponsor whose ultimate goal is to take over the reins of the state. From the proxy's perspective, its military inferiority can influence the desire to collaborate with a senior partner that can offer a degree of protection. We derive the following two hypotheses from this implication for nonstate sponsors:

H1: Nonstate sponsors establish, or partner with, proxies whose identities closely reflect those of the local communities that the sponsor seeks to win over.

H2: Nonstate sponsors select proxies with low military capabilities/expected battlefield utility.

PROXY UTILIZATION. Proxies of state sponsors typically perform active combat roles, including the physical takeover of territory. Based on our earlier theoretical discussion, we expect nonstate sponsors, by contrast, to utilize proxies primarily for security and patrolling purposes in rear areas, or “liberated zones,” rather than for offensive operations. In addition, we expect nonstate sponsors to rely on their proxies as intermediaries with the local population. We derive the following hypotheses from these implications:

H3: Proxies of nonstate sponsors tend to perform security and patrolling operations rather than offensive frontline operations.

H4: Nonstate sponsors rely on proxies for political engagement with the local population.

Closely related to how sponsors utilize their proxies is the question of how the sponsors portray their surrogates. State sponsors are rarely open about their reliance on proxies, frequently disputing their level of control or even denying links to proxies altogether. Our theoretical discussion suggests that nonstate sponsors, by contrast, would have a political interest in acknowledging, and even showcasing, their links to local proxies. We derive the following hypothesis from this implication:

H5: Nonstate sponsors openly acknowledge their ties to proxies.

Finally, relations between state sponsors and proxies are described in the existing literature as asymmetric, with the power balance clearly favoring the sponsor. Nonstate sponsors typically do not match state sponsors in terms of capacity and legitimacy. In addition, our theoretical discussion suggests that in their quest to enhance their legitimacy, nonstate sponsors will find it politically expedient to downplay power asymmetries between themselves and their proxies, while highlighting similarities. We derive the final hypothesis from this implication:

H6: Nonstate sponsors portray their relationship to the proxy as egalitarian rather than hierarchical and seek to obfuscate the differences among themselves and their proxies.

## SELECTION OF CASES

To test the above hypotheses, we examine three case studies of nonstate sponsorship: AQAP's engagement with Sunni tribes in Yemen; the YPG's employment of several proxies under the umbrella of the Syrian Democratic Forces; and Hezbollah's sponsorship of the Lebanese Resistance Brigades. The cases fit our definition of sponsor-proxy relationships while varying in several important respects. First, the nonstate sponsors in these three cases are driven by different ideologies. Hezbollah is a predominantly Shia militant organization; the YPG a mainly leftist revolutionary group with separatist aspirations; and AQAP a militant Sunni jihadist actor. These diverging ideologies indicate that the phenomenon of nonstate sponsors applies to a broad ideological spectrum of armed groups—an important insight considering the lacuna of empirical case studies on nonstate sponsors.

In addition, the case of Hezbollah differs significantly from those of AQAP and the YPG along two additional variables that are of more direct relevance for testing our theoretical argument. First, Hezbollah's military capacity exceeds that of AQAP and YPG by far, as measured by estimates of the size of their respective fighting force, the weapons at their disposal, and their external state support.55 Indeed, Hezbollah is frequently described as a military power-house that is stronger than most of the world's standing armies.56 With a budget estimated at more than 1 billion per year, Hezbollah also surpasses AQAP and YPG in terms of financial resources.57 Second, Hezbollah wields a significant amount of political power in Lebanon and enjoys wide legitimacy even beyond its immediate Shia constituency. In the 2018 parliamentary elections, for example, Hezbollah and its allies won more than 70 of 128 seats, handing the group its biggest electoral success to date, and hence indicating strong popular support for the group within Lebanon.58 Although levels of political influence and legitimacy are hard to quantify, it is of little doubt that AQAP and the YPG are unable to measure up to Hezbollah in this respect as well.59 Hezbollah's attributes as a nonstate actor with unusually high levels of military, financial, and political capacity are significant because, from a methodological standpoint, they allow us to examine the Lebanese organization as the “least likely” case to confirm our theoretical predictions. Hezbollah represents a least likely case because our theory assumes that nonstate sponsors are low-capacity actors that struggle for organizational survival while constrained by gaps in legitimacy and governance. Hezbollah does not fit any of these theoretical assumptions. On the contrary, its military and economic capacity exceeds that of many states, and rather than vying for its survival, Hezbollah is a major political player within Lebanon. It is well known for its provision of social services and governance institutions, and it enjoys a high degree of legitimacy in Lebanon, where its main constituency, the Shia, are the largest minority. Given these attributes, Hezbollah is the least likely of the three cases to sponsor another nonstate actor as a proxy to make further political inroads, and therefore the least likely case to be consistent with our theory's predictions.60 Put differently, evidence in support of our hypotheses in the case of Hezbollah will significantly increase our confidence in the stated theory. ## Case Studies In conducting our case study analyses, we employ a structured focused comparison design.61 Each case consists of two parts, which correspond to the two categories in which we grouped our hypotheses: proxy selection and proxy utilization. We begin by analyzing the case of AQAP, followed by the case of the YPG, and finally the “hard” case of Hezbollah. We conclude the section with a discussion of the findings of our case study analysis. ### AQAP, SUNNI TRIBES, AND THE HADRAMAWT NATIONAL COUNCIL On February 3, 2006, twenty-three jihadi activists, including several commanders of al-Qaida, escaped from a high-security prison in Sanaa. The escapists included Nasir al-Wuhayshi, a former personal secretary to Osama bin Laden who, within a few years, would rebuild a jihadi network on the Arabian Peninsula to become one of al-Qaida's most trusted and important affiliates.62 Although a number of terrorist plots against targets in the West have received particular media attention, most of AQAP's activities focused on Yemen.63 Following the popular uprising against President Ali Abdullah Saleh in 2011 and his formal replacement by Abdrabbuh Mansour Hadi in 2012, the group stepped up its local insurgency campaign. In doing so, AQAP heeded the strategic guidance of bin Laden who, in a letter penned around mid-2010 to his then chief of staff, Attiyah abd al-Rahman, advised that Yemen was not ripe for a jihadi takeover through a strategy focused on violence. Instead, he recommended that, in Yemen, jihadis adopt a strategy focused on media operations, gradual preparation of the conditions for a jihadi takeover, and da'wa efforts—the latter indicating a nonviolent form of outreach aimed at the gradual mobilization of Muslims for the jihadi cause.64 AQAP PROXY SELECTION. A crucial element in this strategy was bin Laden's directive to refrain from attacking the tribes, and instead seek ways to engage them. The al-Qaida leader believed that the tribes were naturally disposed to supporting the jihadi project and argued that historically, whenever the mujahideen treated the tribes respectfully, they would support the jihadis in return.65 Heeding bin Laden's general advice and agreeing that the tribes were the true power brokers in Yemen, then-AQAP leader Wuhayshi described the tribes as the greatest hurdles toward AQAP's desire to control territory. Control, Wuhayshi realized, was contingent on successful engagement with the tribal leaders. He also understood that a main challenge was that many tribes were reluctant to cede control, especially to outsiders.66 To try to overcome this problem, AQAP engaged in a rebranding effort that would be mindful of tribal norms, customs, and preferences. This meant that AQAP would henceforth underplay its ties to external organizations. To that end, it established a locally focused parallel movement, Ansar al-Sharia.67 Under the movement's brand, the group managed to expand its hold over several towns in the southern Yemeni provinces of Abyan and Shabwa. Initially, AQAP's attempt to leverage tribal support was short-lived. In al-Bayda, the group meddled in an intra-tribal feud by backing Tariq al-Dhahab, the son of a prominent leader of the Qaifah tribe who felt he had inherited less land and wealth from his deceased father than was his due.68 However, AQAP made the mistake of doing so in an overt manner and insisting that al-Dhahab's fighters proclaim their loyalty to the group, causing backlash among the tribes and eventually leading to AQAP's expulsion from al-Bayda.69 AQAP was quick to learn the appropriate lessons from its initial foray into Bedouin tribal politics and henceforth adopted a more loose and pragmatic approach to building partnerships. It rested on the distribution of funds to tribal elements willing to fight as surrogates.70 This pragmatic stance helped AQAP regain control over some provinces, including al-Bayda, and over southern parts of Hadramawt. AQAP employed this approach when it captured the port city of al-Mukalla—the fifth largest city in Yemen—in April 2015, benefiting from the unfolding crisis of governance with Iran-backed Houthi rebels taking over the presidential palace in the capital of Sanaa earlier that year. In al-Mukalla, AQAP cooperated with local tribes to form the Hadramawt National Council (HNC), a proxy militia tasked with helping secure critical infrastructure and provide basic services such as drinking water, electricity, and fuel to the population.71 The Council was reportedly led by prominent local tribal elders and religious scholars and oversaw approximately 5,000 militia members.72 According to HNC officials, the Council had received funding from AQAP but lacked weapons, ammunition, and salaries for its “security officers” even though AQAP had promised to provide them with light arms for self-defense.73 Thus, despite meeting or even exceeding AQAP numerically, the HNC was no match in terms of military capabilities.74 AQAP PROXY UTILIZATION. Almost immediately after AQAP had captured al-Mukalla, it handed over government institutions to the newly founded HNC and provided it the equivalent of several million U.S. dollars of funding—money it seized from al-Mukalla's central bank—to provide basic services.75 Besides offering protection to key facilities such as government institutions and the local airport, the HNC apparently focused on the electricity sector, maintenance of the seaport, and other services such as health care, street cleaning, and water services.76 To prepare its security force, the Council created training centers in several districts of Hadramawt.77 Apparently, the HNC's other important function was to act as intermediary between AQAP and the Yemeni government. HNC Secretary-General Abdul-Hakeem bin Mahfood explicitly acknowledged this role in an interview with al-Jazeera and claimed that his organization had also conducted talks with the Saudi government.78 Delegating day-to-day governance to the HNC allowed AQAP to focus on recruitment, training, and other crucial aspects of its insurgency campaign.79 AQAP has aimed to replicate this strategy, known as al-yad al-makhfi (invisible hand), in other places. It has tried to insert cells within tribal militias and other nominally pro-government forces that are fighting the Houthi rebels, offering them financial and military aid.80 Before taking over al-Mukalla, AQAP adopted the moniker “Sons of Hadramawt,” which it must have considered more appealing to the local population than the al-Qaida label. It also made sure not to raise any jihadist flags when it captured the city.81 AQAP deliberately blurred the lines between formal members and sympathizers, thereby trying to lower the bar of support for AQAP to include a wider spectrum of the local population. No longer was it necessary to swear an oath of allegiance to al-Qaida to do the group's bidding. As one AQAP commander told an International Crisis Group (ICG) researcher, “We are as one with the [Sunni] tribes like never before. We are not al-Qaida now. Together we are the Sunni army.”82 The group also made sure to exploit the sectarian divide that separated the Sunni tribes from the central government that, after the Houthi takeover, was dominated by Zaidi Shia and widely believed to be supported by Iran. AQAP missed few opportunities to highlight the sectarian divide in the country, but simultaneously continued to downplay its al-Qaida ties in an attempt to present itself as the defender of all Sunni Muslims vis-à-vis Shia expansionism. AQAP also made sure to use specific narratives attuned to Yemeni tribal tradition and customs in its propaganda, rather than relying on generic religious tenets.83 AQAP SUMMARY. AQAP's control of al-Mukalla was a relatively short-lived affair—the group withdrew before Emirati-backed forces liberated the city in April 2016. Nevertheless, controlling a city even briefly provided AQAP with the experience to govern by proxy and build relations with the local elite.84 Its efforts to engage with the population were met with at least partial success, with several locals claiming that life in al-Mukalla has worsened since AQAP withdrew.85 According to Horton, AQAP's rule over al-Mukalla “allowed it to establish its reputation as a reliable and relatively capable force that was willing to work with those elites whose interests overlapped with its own,” thereby creating limited local support that was central to ensure its long-term survival.86 In other instances, however, relations between AQAP and the tribes were uneven and, in many cases, failed to endure. Engagement with the tribes appeared to work better in areas where the tribal structure had been weaker to begin with. In areas that saw a stronger preexisting tribal structure, AQAP had a more limited presence.87 ### THE PEOPLE'S PROTECTION UNITS AND THE SYRIAN DEMOCRATIC FORCES As the Islamic State grabbed global attention with its lightning advances in Iraq and Syria starting in 2014, the U.S.-led coalition was desperate to find reliable partners on the ground to push back against the jihadi insurgents. In Syria, starting with the Islamic State's siege of Kobane in September 2014, the United States embarked on a partnership with the People's Protection Units, the armed wing of the Kurdish Democratic Union Party (PYD).88 While the Kurdish militia proved exceptionally capable of slowing down, halting, and eventually rolling back the Islamic State, it also caused unprecedented tensions between the United States and Turkey.89 The latter considers the PYD as an extension of the Kurdistan Workers’ Party (PKK), which has waged an insurgency against Turkey for several decades. Indeed, there is little dispute that the PKK and the PYD are closely intertwined and that the latter adopted PKK leader Abdullah Öcalan's ideological tenets.90 YPG PROXY SELECTION. When the Barack Obama administration decided that the United States should collaborate with the YPG in the fight against the Islamic State starting in the fall of 2014, the ties between the Syrian Kurds and the PKK presented Washington with a dilemma: both the United States and the European Union consider the latter a terrorist organization.91 At first, the United States utilized Iraqi Peshmerga militias to arrange for indirect weapons deliveries to the YPG and its all-female unit, the Women Protection Units (YPJ).92 The Obama administration then encouraged the establishment of the Syrian Democratic Forces (SDF), formed in October 2015 as a nominal coalition of Kurdish, Sunni Arab, Turkmen, and Syriac Christian militias.93 A closer look, however, reveals that the YPG is clearly the dominant force.94 According to the ICG, “the YPG is in overall command and controls military supply,” and its “commanders are in charge of military logistics, providing the most powerful weapons to YPG fighters.”95 Some of the smaller SDF groups’ ties to the YPG preceded the coalition's emergence. The Syriac Military Council (MFS), for example, was officially established in January 2013 and declared its intention to join the YPG one year later.96 In reality, however, the MFS had been “working under the YPG,” as members have admitted in interviews.97 The militia continues to wear distinct insignias.98 Another SDF unit, Jabhat al-Akrad (Kurdish Front), emerged in 2013 as a part of the Free Syrian Army. Given its links to the Free Syrian Army and its inclusion of Arab fighters, Wladimir van Wilgenburg, an expert on the Kurds, assesses that Jabhat al-Akrad was most likely created by the PYD “to gain access to mixed Arab-Kurdish areas and to make logistics between the three Kurdish enclaves easier.”99 In other cases, the SDF created new units, arguably in a bid to obscure the YPG's dominant role. For example, in April 2016, prior to the battle for Manbij, the SDF set up the so-called Manbij Military Council (MMC), which consisted of mainly Arab and Turkmen fighters and was supposed to govern the city upon its liberation from the Islamic State.100 YPG PROXY UTILIZATION. Non-Kurdish elements of the SDF in general are small and do not have any meaningful impact on the battlefield. Instead, Arab SDF fighters take on secondary roles such as maintaining local security.101 Similarly, the Syriac MFS also appears to avoid major military battles and focus on patrols.102 For the YPG, however, the reason for including smaller non-Kurdish groups lies in additional political influence vis-à-vis their surrogates’ respective communities, while providing the SDF with political cover, namely the appearance of being a truly diverse and multiethnic entity.103 While the military strength of the Arab, Turkmen, and Syriac units of the SDF remained negligible, their mere presence increased the YPG's freedom of operations and enhanced its legitimacy as it began to advance beyond Kurdish areas and move deeper into Islamic State–controlled territory.104 An MFS spokesperson explicitly acknowledged this, explaining that although the MFS is “not the dominant force in military terms,” it could provide a link between Kurds and Arabs.105 From the YPG's perspective, one of the SDF's major purposes is to assuage the concerns and distrust of Sunni Arabs living in areas such as Manbij or Tel Abyad toward the Kurds.106 Hence, when the SDF took Manbij, the YPG made up the bulk of the fighting forces that captured the city before ostensibly handing it over to the MMC.107 MMC spokesperson Shervan Derwish acknowledged in a New York Times oped that it was “the Kurdish fighters who liberated Manbij,” whereas he described the MMC as “a security force composed primarily of local Arabs, to hunt down terrorists and sleeper cells.”108 Publicly, the YPG claims that the other SDF components are equal partners. As two analysts put it, “SDF promotional material tells the story of a multiethnic band that strives together towards decentralized, participatory government, which would protect the rights of all ethnic groups without emphasizing their distinct identities.”109 In reality, such pretensions are little more than a charade, according to an ICG report describing the situation in Manbij.110 Even after officially withdrawing a couple of months later, the YPG made sure to retain groups of military advisers whose task was to train and thus, by extension, control the MMC.111 The YPG has also made sure that the other groups fighting within the SDF's ranks are receptive toward the ultimate Kurdish vision and desired outcome. As reported by the Washington Post, before receiving military training by U.S. special forces, new Arab recruits are required to learn and embrace PKK leader Öcalan's ideological tenets.112 YPG SUMMARY. Although smaller non-Kurdish units may have hoped that aligning with the SDF and the YPG would provide them with international backing, thereby enhancing their relevance, survivability, and ability to reap benefits from their participation in the Syrian conflict, it has been the YPG that has benefited from the SDF more than any other group. Not only did the SDF prove an ideal conduit to receive weaponry and other equipment from international backers, but it also allowed the YPG to collaborate with some of the world's most powerful armed forces, effectively protecting it from air strikes by Russia or the regime of Syria's president, Bashar al-Assad.113 Even more importantly, under the guise of this supposed multiethnic and equal partnership, the YPG has managed to expand its control beyond its core territories, including capturing Syria's largest oil field.114 ### HEZBOLLAH AND THE LEBANESE RESISTANCE BRIGADES Scholars typically consider Hezbollah as Iran's proxy of choice.115 The “Party of God” is responsible for the 1983 bombings of U.S. Marines and French paratroopers facilities in Beirut, the kidnappings of Western diplomats in Lebanon during the 1980s, and terrorist attacks in 1992 and 1994 against Israeli and Jewish targets in Argentina on behalf of Iran.116 Under the direction of the Iranian Quds Force and following the U.S. invasion in 2003, Hezbollah has also operated in Iraq, and later in Yemen, and in Syria, where it proved crucial to the Assad regime's survival.117 In these conflict theaters, Hezbollah has adopted a sponsorship role of other armed nonstate groups. Recently, it has trained and assisted Shia militias in Iraq in their fight against the Islamic State.118 It has offered similar support to some Syrian pro-government militias, effectively using them as surrogates.119 In those cases, however, Hezbollah acted primarily at the behest of Tehran.120 A more clear-cut case of Hezbollah establishing a proxy in its own right is the Lebanese Resistance Brigades, a paramilitary group consisting of Christians, Sunnis, Shia, and Druze. HEZBOLLAH PROXY SELECTION. In the fall of 1997, Hezbollah Secretary-General Hassan Nasrallah announced the creation of the Lebanese Brigades for Resisting the Israeli Occupation (Saraya al-Muqawama al-Lubnaniya, aka Saraya).121 Nasrallah proclaimed that the unit would be open to all volunteers regardless of their religion. Its members were to receive training from regular Hezbollah operatives and would then deploy against Israeli soldiers in the south.122 According to Nicholas Blanford, the group published ads in Lebanese newspapers, including a phone number that potential volunteers could call in order to receive more information on how to join the group.123 Recruitment was conditional upon two main requirements, namely to fulfill certain physical demands and to have no links to Israel or its allies.124 Hezbollah made sure to emphasize that the Brigades were open to any Lebanese wanting to join its fight against Israel. During a press trip to a Saraya training camp in November 1999, a Hezbollah fighter accompanying the participating reporters made sure to point out the multiethnic and multireligious composition of the Brigades consisting of Sunni, Christian, Shia, Druze, and even Armenian members.125 After Israel retreated from Lebanon in May 2000, the Brigades were disbanded, but Hezbollah chose to revive them in the aftermath of the Second Lebanon War in 2006.126 Apparently, Hezbollah provides weapons to Saraya members and pays them money, although they do not seem to receive a monthly salary like regular Hezbollah fighters.127 In some cases, Hezbollah appears to have provided heavy weaponry, including Russian anti-tank missiles, to a separate all-Christian unit within the Brigades that changed its name to the Self-Defence Committee and carried out armed patrols along the Syrian-Lebanese border adjacent to the Beqaa Valley.128 HEZBOLLAH PROXY UTILIZATION. Following their inception, the Resistance Brigades initially launched more than 150 attacks against Israel and its South Lebanese Army proxy, including small-unit ambushes, mortar attacks, and the use of land mines.129 These assaults amounted to a mere fraction of those carried out by Hezbollah or Amal. Moreover, some observers questioned whether the Saraya was responsible, or whether Hezbollah's media department simply let it take credit.130 Scrutinizing the modest military capabilities and effectiveness of the Saraya, however, is to misunderstand its main purpose. Rather than to augment Hezbollah's operational capacity, the creation of the Saraya should be considered as a bid by Hezbollah to broaden its appeal to Lebanese of all denominations. Upon the creation of the Brigades, Nasrallah himself had declared that it was a response to the growing number of Lebanese wishing to “join the resistance.”131 As Chris Zambelis argues, the Saraya can be seen as part of Hezbollah's “progression toward political expediency and pragmatism after its formal entry into Lebanese politics,” commonly referred to as the group's “Lebanonization.”132 Following its revival after the 2006 war, the Saraya once again served to boost Hezbollah's national and patriotic credentials and build an “esprit de corps within the parliamentary opposition to Fouad Siniora's government between 2006 and 2008” by providing training to Hezbollah's political allies, including Christians, Sunnis, and Druze.133 At the same time, it was obvious that Hezbollah did not expect it to play a significant combat role in future confrontations, but rather to “look after the refugees from the south.”134 More recently, Saraya members have reportedly deployed to Syria alongside Hezbollah.135 They do not seem to fight on the front lines, however, instead providing logistical support.136 Even more important is their symbolic role, acting as intermediaries for Hezbollah in Sunni and Christian areas and thus increasing its freedom of operation.137 At the same time, the majority of Saraya's members remain in Lebanon for patrol duties.138 Hezbollah and Saraya members acknowledge differences in their ideological training, but claim that both receive the same military training, even though analysts claim that the Brigades’ military education is less extensive.139 The latter would confirm assessments that there is clear separation between members of Hezbollah and members of the Resistance Brigades to “prevent any penetration of the Islamic Resistance.”140 The Saraya were thus also an effort to attract additional recruits while maintaining operational security. To reinforce this distinction, members of the Brigades wear different uniforms, although this might differ during deployments in Syria where they apparently use Hezbollah insignia; additionally, the Brigades has its own flag.141 HEZBOLLAH SUMMARY. The Lebanese Resistance Brigades has never had a significant impact on Hezbollah's military capabilities. Rather, the Saraya has to be seen as part of Hezbollah's efforts to appeal to Lebanese beyond its immediate Shia constituency, usually described as the so-called phase of al-infitah, the opening.142 Moreover, in recent years Hezbollah has used the Saraya in an effort to rebrand itself as a self-proclaimed “bulwark against violent Islamist extremism and sectarianism in Lebanon […] position[ing] itself as a unifying and even moderate stabilizing force in light of what is a catastrophic alternative promoted, according to Nasrallah, by the likes of Saudi Arabia and other Persian Gulf Arab monarchies.”143 ### CASE STUDY FINDINGS Our analysis shows strong support for our hypotheses in the cases of AQAP and the YPG and moderate support in the “hard case” of Hezbollah (see table 2). Table 2. Summary of Case Study Analysis Nonstate Sponsor IssueHypothesisAl-Qaida in the Arabian PeninsulaPeoples’ Protection UnitsHezbollah Proxy Selection H1: Nonstate sponsors (NSSs) partner with proxies whose identities closely reflect those of the local communities. strong support strong support strong support H2: NSSs are militarily superior to their proxies. strong support strong support strong support Proxy Utilization H3: Proxies of NSSs tend to perform security and patrolling operations strong support strong support moderate support H4: NSSs rely on proxies for political engagement with locals. strong support strong support strong support H5: NSSs openly acknowledge ties to their proxies. strong support strong support strong support H6: NSSs highlight relationships as egalitarian, gloss over differences with proxies. strong support moderate support weak support Nonstate Sponsor IssueHypothesisAl-Qaida in the Arabian PeninsulaPeoples’ Protection UnitsHezbollah Proxy Selection H1: Nonstate sponsors (NSSs) partner with proxies whose identities closely reflect those of the local communities. strong support strong support strong support H2: NSSs are militarily superior to their proxies. strong support strong support strong support Proxy Utilization H3: Proxies of NSSs tend to perform security and patrolling operations strong support strong support moderate support H4: NSSs rely on proxies for political engagement with locals. strong support strong support strong support H5: NSSs openly acknowledge ties to their proxies. strong support strong support strong support H6: NSSs highlight relationships as egalitarian, gloss over differences with proxies. strong support moderate support weak support H1, which hypothesizes that nonstate sponsors establish, or partner with, proxies whose identities closely reflect those of the local communities that the sponsor seeks to win over, was strongly supported in all three case studies. In the case of AQAP, the group's decision to engage with Sunni Bedouin tribes in Yemen stemmed from the recognition—expressed by al-Qaida leader Osama bin Laden personally and confirmed by then leader of AQAP, Nasir al-Wuhayshi—that the tribes were the main power brokers in Yemen. In the case of the YPG, groups such as the Syriac Military Council and the Manbij Military Council included non-Kurdish fighters, suggesting that their ethnic composition played a role in their selection as proxies. In the case of Hezbollah's sponsorship of the Saraya, Hezbollah leaders went out of their way to emphasize that the group was open to Lebanese of all ethnic and denominational backgrounds. All three cases also strongly support H2, which hypothesizes that nonstate sponsors are militarily superior to their proxies. In the case of AQAP, this was evident in the group's withholding of arms and ammunition, and its general proclivity to maintain control over military aspects such as recruitment and training, even if the Hadramawt National Council, AQAP's main proxy partner, outnumbered the AQAP. In the case of the YPG, the group maintained military primacy over its proxies, retained control over supplies and logistics, and dominated the battlefield with its numerically superior forces. In the case of Hezbollah, the group did provide some weapons to its proxy, the Saraya, but the latter is clearly no match to its sponsor given the latter's superior military capacity. In sum, the strong support for H1 and H2 combined lends strong credence to the idea that nonstate sponsors select proxies based not on their military capability, but on the perceived political advantages that they afford to their sponsors. Whereas H1 and H2 address the issue of proxy selection, the remaining hypotheses relate to the sponsors’ proxy utilization. Here, the cases of AQAP and the YPG provide, respectively, very strong and strong support for H3, H4, H5, and H6. The case of Hezbollah offers partial confirmation of our hypotheses. H3, which hypothesizes that proxies of nonstate sponsors tend to perform security and patrolling operations rather than offensive frontline operations, is strongly supported by the cases of AQAP and YPG, and moderately supported by the case of Hezbollah. AQAP provides evidence through its cooperation with the HNC, which was tasked with providing basic services to the population. In the case of the YPG, the group relies on Jabhat al-Akrad for logistical support and on the MMC for governance following the liberation of Manbi. It also used Arab SDF fighters for secondary roles such as the provision of security.144 In the case of Hezbollah, the evidence is less conclusive. Although the Resistance Brigades reportedly carried out attacks against Israel and its Lebanese allies, several analysts wondered whether Hezbollah was actually responsible for the attacks, allowing its proxy to take credit. At a later point, the Saraya appears to have played a mostly logistical role. And although some Saraya members were dispatched to Syria, most remained in Lebanon to perform patrol functions. All three cases also strongly support H4, which hypothesizes that nonstate sponsors rely on proxies for political engagement with the local population. Thus, AQAP used HNC as an intermediary between al-Qaida and both the Yemeni and Saudi governments, as a senior HNC official admitted. As concerns the YPG, a spokesman for the MFS acknowledged the group's secondary military role, but emphasized that his group was a link between Kurds and Arabs. This link became of key importance when the SDF repelled the Islamic State from territories not previously controlled by Kurds. In the case of Hezbollah, the case study shows that the Saraya provided military training to Hezbollah's non-Shia partners, while also acting as intermediaries for Hezbollah in Sunni and Christian areas. The cases also support H5, which hypothesizes that nonstate sponsors openly acknowledge their ties to their sponsor. In the case of AQAP, this acknowledgment was reflected in a statement by an AQAP commander that his group is “as one with the tribes.”145 In the case of the YPG, the group's ties to its proxies are publicly acknowledged in as far as both the YPG and its proxies are a part of the SDF umbrella group. In the case of Hezbollah, the group's ties to the Saraya are also openly acknowledged, with Hezbollah having made public appeals to the Lebanese population to join the Resistance Brigades. Finally, the cases also offer moderate support to H6, which hypothesizes that nonstate sponsors seek to frame their relationships to their proxies as egalitarian, while glossing over their differences. The case of AQAP provides the strongest support for H6. AQAP downplayed its ties to outside organizations (including the al-Qaida leadership) and went so far as establishing a parallel front movement, Ansar al-Sharia, to obfuscate its ties to al-Qaida—a foreign group—while portraying itself as more organic to Yemen. To that end, it also exempted the Bedouin tribes from swearing an oath of allegiance and largely refrained from raising AQAP flags when capturing Yemeni cities. AQAP commanders exploited sectarian divisions as a way to further strengthen the bonds with its proxy partners based on their shared Sunni identity. Regarding the YPG, major capacity gaps exist between the group and its proxies. Nevertheless, the case study shows that the YPG made at least some attempts at promoting the idea that all the groups within the SDF were equal partners; among these efforts was the use of promotional material. As an ICG report stated, the YPG's “official rhetoric signals inclusiveness and pluralism,” even if the YPG dominates its partners in practice.146 The case of Hezbollah is similar to the YPG in this regard. While Hezbollah, in accordance with its “Lebanonization” efforts, pays lip service to the idea of a broad-based resistance that transcends sectarian boundaries, the group retains a clear separation between itself and its proxy, not least for reasons of operational security. ## Conclusion We have argued that nonstate sponsors rely on proxies primarily as political ancillaries rather than military surrogates. As such, they offer a value to their sponsors that is different from that which proxies provide to state sponsors. These proxies are more representative of certain segments of the local population than are their nonstate sponsors, thus granting their senior partners access in environments that are often less than welcoming. They enhance their sponsors’ freedom of operations, but do so by political means such as mediation and negotiation in specific theaters rather than through offensive military efforts. They are important, and sometimes critical assets that enhance the legitimacy, and hence the political aspirations, of their sponsor. At the same time, the analysis of our three case studies suggests some qualifications to our argument. The cases provide stronger support for the hypotheses related to proxy selection than for those related to how sponsors utilize and portray their proxies. The notion that nonstate sponsors select proxies based on their perceived political advantages is strongly supported in all three case studies. The three nonstate sponsors that we examined are less uniform, however, in the way that they utilize and relate to their proxies. While the case of AQAP best fits our theoretical predictions, the examination of Hezbollah reveals some limitations to our argument. Contrary to our expectations, Hezbollah appears to have employed the Saraya in part for offensive operations, and it treats its proxy as a junior partner, despite some half-hearted claims to the contrary. Neither does it seem to employ its proxy for governance tasks—contrary to AQAP, which has handed most of these matters to its Bedouin partners. This finding suggests that our theoretical predictions apply better to nonstate actors with low and moderate capacity levels, such as AQAP and YPG, than to high-capacity nonstate sponsors such as Hezbollah. In retrospect, this finding is of little surprise. Actors with modest capacities are more dependent on political support, given their more precarious situational context. They can hardly afford to alienate their proxies, which are quasi representatives of the local population on whose support the sponsor depends. In contrast, a group such as Hezbollah is not critically dependent on proxy forces, even if it welcomes the prestige boost that such a relationship affords. Given Hezbollah's excessive military power; its political strength inside Lebanon; and its organizational capacity, which obviates the need to delegate the role of governance services to partners, Hezbollah is immune to threats of defection on the part of its proxy.147 Consequently, it will naturally be more controlling—much like state sponsors approach the relationship with their proxies. In practice, then, high-capacity actors such as Hezbollah fall in between the expected behavior of an ideal-typical nonstate sponsor and that of an ideal-typical state sponsor. More generally, our findings imply that the greater the capacity and legitimacy of a nonstate sponsor, the more it will resemble a state sponsor in its behavior. At the same time, a nonstate sponsor such as Hezbollah is clearly an outlier among nonstate actors, most of which possess more modest capacities. Consequently, we believe that most nonstate sponsors will utilize proxies in a way that is consistent with our theoretical expectations. The findings of our study also have implications for contemporary sponsor-proxy relationships more broadly. Assuming, as we do, that most nonstate sponsors are low-capacity rather than high-capacity sponsors, the phenomenon of nonstate sponsorship suggests that the asymmetries in power and capability between nonstate sponsors and proxies are not as pronounced as those in traditional cases of state sponsorship. Compared with the asymmetries between traditional state sponsors and their proxies, the “capability gap” between nonstate sponsors and their proxies is narrower, which has repercussions for the duration and stability that one can expect of those sponsor-proxy relationships that fit the pattern described in our study. Power gaps between nonstate sponsors and their proxies are narrow for at least two reasons. First, most nonstate actors—including those that are sponsors—possess relatively low material and economic capacity to begin with, which places natural limits on the magnitude of the asymmetry of capabilities between themselves and their proxies. Second, proxies of nonstate sponsors offer unique advantages that help further reduce the capability gap between sponsors and proxies. Their outsized importance levels the playing field between proxies and nonstate sponsors, rendering these relationships less lopsided than traditional state-centric sponsor-proxy relations. These dynamics can also affect the division of labor between sponsors and proxies. Although sponsors are likely to continue, for the most part, to provide such traditional services as funding, weapons, training, and advisory or intelligence support, the future role of proxies may no longer be limited to the provision of manpower, as has traditionally been the case when the sponsors were state actors.148 In cases of nonstate sponsorship, proxies primarily provide services such as political legitimacy, which, in the conventional understanding of proxy relationships, are rather atypical—or at the very least no more than a secondary concern. But even where proxies provide services to their nonstate sponsor that better fit the more conventional understanding of proxies—for example, superior knowledge of terrain—such services are of far more critical importance to nonstate sponsors than they would be to state sponsors. As a result, the overall value that proxies of nonstate sponsors provide to such relationships is greater when compared to the value that proxies of states provide to their benefactors. Proxy relationships will continue to bring together sponsors and proxies that will divide their labor based on the principle of comparative advantage. But as the capability gap between the parties narrows, it will not always be evident which actor holds which relative advantage. The narrowing capability gap described above is likely to produce relations that are less consistent and durable than traditional sponsor-proxy relationships involving state actors. Nonstate actors adopting sponsorship roles will often have to team up with local actors whose interests in maintaining such arrangements are inconsistent at best, and unpredictable at worst. As the case studies showed, the proxies of nonstate sponsors are oftentimes tribes or ad hoc “popular committees” made up of locals with highly particularistic interests. These micro-actors are less beholden to static ideologies and far more likely to act in pursuit of more mundane self-interest. Their ties to sponsors will often be uneven and fluctuating, as they are likely to sell their services to the highest bidder. As nonstate actors are playing an ever more central role in international affairs, it is likely that their role as sponsors will grow accordingly.149 Sponsor-proxy relations are therefore likely to become more common in the future. Such relations will most likely be dynamic, adaptive, and pragmatic, with the partners willing to shift their loyalties on short notice. The resulting relationships will be less stable, predictable, and enduring, and instead more inconsistent and transactional over time. The implication is twofold. On the one hand, more, but also more inconsistent, sponsor-proxy constellations are likely to emerge in the foreseeable future. This suggests that future armed conflicts will grow even more complex and intractable. On the other hand, the transactional nature and pragmatism of such actors may suggest that even the most dangerous sponsor-proxy relationships face considerable hurdles to their long-term survival. Such relationships will stand on shaky grounds and will be susceptible to pressures and manipulations induced both by their partners and by external actors intent on undermining these relationships. ## Acknowledgments The authors acknowledge the support provided by the Center on National Security at Fordham Law School's Proxy Violence in the Middle East and Beyond project in partnership with the Carnegie Corporation of New York. For their helpful comments and suggestions, the authors thank the anonymous reviewers and participants in seminars and conferences where this research was presented: the Center for National Security at Fordham Law; the Lauder School of Government at the Interdisciplinary Center Herzliya; and annual conventions of the International Studies Association and the American Political Science Association. ## Notes 1. Andrew Mumford, Proxy Warfare (Cambridge: Polity, 2013), p. 11; Geraint Hughes, My Enemy's Enemy: Proxy Warfare in International Politics (Eastbourne, U.K.: Sussex Academic Press, 2012), p. 11; and Tyrone L. Groh, Proxy War: The Least Bad Option (Stanford, Calif.: Stanford University Press, 2019), pp. 28–29. 2. For historical examples, see Thucydides, The Peloponnesian War, trans. Martin Hammond (Oxford: Oxford University Press, 2009); Edward N. Luttwak, The Grand Strategy of the Roman Empire: From the First Century A.D. to the Third (Baltimore, Md.: Johns Hopkins University Press, 1976); Charles J. Esdaile, Fighting Napoleon: Guerrillas, Bandits, and Adventurers in Spain, 1808–1814 (New Haven, Conn.: Yale University Press, 2004); Andreas Krieg and Jean-Marc Rickli, Surrogate Warfare: The Transformation of War in the Twenty-First Century (Washington, D.C.: Georgetown University Press, 2019); Alex Marshall, “From Civil War to Proxy War: Past History and Current Dilemmas,” Small Wars & Insurgencies, Vol. 27, No. 2 (April 2016), pp. 183–195, doi.org/10.1080/09592318.2015.1129172; and Geraint Hughes, “Militias in Internal Warfare: From the Colonial Era to the Contemporary Middle East,” Small Wars & Insurgencies, Vol. 27, No. 2 (April 2016), pp. 196–225, org/10.1080/09592318.2015.1129171. 3. Chris Loveman, “Assessing the Phenomenon of Proxy Intervention,” Conflict, Security & Development, Vol. 2, No. 3 (December 2002), pp. 29–48, doi.org/10.1080/14678800200590618; Michael A. Innes, ed., Making Sense of Proxy Wars: States, Surrogates, and the Use of Force (Washington, D.C.: Potomac, 2012); Hughes, My Enemy's Enemy; Mumford, Proxy Warfare; Groh, Proxy War; Eli Berman and David A. Lake, eds., Proxy Wars: Suppressing Violence through Local Agents (Ithaca, N.Y.: Cornell University Press, 2019); and Candace Rondeaux and David Sterman, “Twenty-First Century Proxy Warfare: Confronting Strategic Innovation in a Multipolar World since the 2011 NATO Intervention” (Washington, D.C.: New America Foundation, February 2019). 4. During the Cold War, it was also common to view less powerful states as proxies, as exemplified in the Soviet sponsorship of Cuba. For a critical appraisal of this particular relationship, see Piero Gleijeses, “Moscow's Proxy? Cuba and Africa, 1975–1988,” Journal of Cold War Studies, Vol. 8, No. 4 (Fall 2006), pp. 98–146, org/10.1162/jcws.2006.8.4.98. 5. Daniel Byman et al., Trends in Outside Support for Insurgent Movements (Santa Monica, Calif.: RAND Corporation, 2001), pp. 71–78; Mumford, Proxy Warfare, p. 45; Christopher Phillips and Morten Valbj⊘rn, “‘What Is in a Name?': The Role of (Different) Identities in the Multiple Proxy Wars in Syria,” Small Wars & Insurgencies, Vol. 29, No. 3 (May 2018), pp. 414–433, doi.org/10.1080/09592318.2018.1455328; and Rondeaux and Sterman, “Twenty-First Century Proxy Warfare,” pp. 50–51. 6. According to data collected by David E. Cunningham, Kristian Skrede Gleditsch, and Idean Salehyan, for example, external actors provided explicit or alleged support to 48 percent of 443 rebel groups engaged in armed conflict from 1945 to 2011. See Cunningham, Gleditsch, and Salehyan, “Non-State Actors in Civil Wars: A New Dataset,” Conflict Management & Peace Science, Vol. 30, No. 5 (November 2013), p. 527, org/10.1177%2F0738894213499673. 7. Kenneth A. Schultz, “The Enforcement Problem in Coercive Bargaining: Interstate Conflict over Rebel Support in Civil Wars,” International Organization, Vol. 64, No. 2 (Spring 2010), pp. 281–312, doi.org/10.1017/S0020818310000032; and Erica Dreyfus Borghard, “Friends with Benefits? Power and Influence in Proxy Warfare,” Ph.D. dissertation, Columbia University, 2014, pp. 4–5. 8. Colin P. Clarke, After the Caliphate: The Islamic State and the Future of the Terrorist Diaspora (Cambridge: Polity, 2019). 9. Military and political objectives are clearly related, with military goals being designed to achieve broader political objectives. Nevertheless, we argue that these different actor types place greater relative weight on one strategy over another. 10. There is some variation between democratic and authoritarian regime types, with the latter being more willing to wage war. Compared to nonstate actors, however, the basic argument holds true. For variation in regime type and war proneness, see David A. Lake, “Powerful Pacifists: Democratic States and War,” American Political Science Review, Vol. 86, No. 1 (March 1992), pp. 24– 37, doi.org/10.2307/1964013; Joe D. Hagan, “Domestic Political Systems and War Proneness,” Mershon International Studies Review, Vol. 38, No. 2 (October 1994), pp. 183–207, doi.org/10.2307/222714; and Bruce Bueno de Mesquita and Randolph M. Siverson, “War and the Survival of Political Leaders: A Comparative Study of Regime Types and Political Accountability,” American Political Science Review, Vol. 89, No. 4 (December 1995), pp. 841–855, org/10.2307/2082512. 11. Philip Towle, “The Strategy of War by Proxy,” RUSI Journal, Vol. 126, No. 1 (1981), p. 21, doi.org/10.1080/03071848108523403; Mumford, Proxy Warfare, pp. 11-12, 26-29; Marshall, “From Civil War to Proxy War”; Hughes, “Militias in Internal Warfare”; and Rondeaux and Sterman, “Twenty-First Century Proxy Warfare,” pp. 18-19. 12. Vladimir Rauta, “A Structural-Relational Analysis of Party Dynamics in Proxy Wars,” International Relations, Vol. 32, No. 4 (December 2018), p. 450, org/10.1177%2F0047117818802436. 13. Ibid., pp. 452–457. 14. See Michael G. Findley and Tze Kwang Teo, “Rethinking Third-Party Interventions into Civil Wars: An Actor-Centric Approach,” Journal of Politics, Vol. 68, No. 4 (November 2006), pp. 828-837, org/10.1111/j.1468-2508.2006.00473.x. 15. The common reliance on proxy wars or proxy conflict is problematic because certain relationships between sponsors and proxies—for example, the relationship between Iran and Hezbollah— stretch well beyond a single theater of conflict. In addition, the implication that the use of proxies is the dominant strategic characteristic of these conflicts is unsubstantiated. 16. This definition builds on previous works on proxy relationships. See Idean Salehyan, “The Delegation of War to Rebel Organizations,” Journal of Conflict Resolution, Vol. 54, No. 3 (June 2010), p. 503, doi.org/10.1177%2F0022002709357890; Hughes, My Enemy's Enemy, p. 11; Mumford, Proxy Warfare, p. 11; Rauta, “A Structural-Relational Analysis of Party Dynamics in Proxy Wars,” p. 457; Groh, Proxy War, p. 29; and, especially, Borghard, “Friends with Benefits?” p. 17. 17. Salehyan, “The Delegation of War to Rebel Organizations,” pp. 503–504; Hughes, My Enemy's Enemy, pp. 27–29; and Groh, Proxy War, pp. 35–37. 18. Glenn H. Snyder, Alliance Politics (Ithaca, N.Y.: Cornell University Press, 1997), p. 4. See also Stephen M. Walt, The Origins of Alliances (Ithaca, N.Y.: Cornell University Press, 1987). 19. Ely Karmon, Coalitions between Terrorist Organizations: Revolutionaries, Nationalists, and Islamists (Leiden, Netherlands: Brill, 2005); Fotini Christia, Alliance Formation in Civil Wars (New York: Cambridge University Press, 2012); Assaf Moghadam, Nexus of Global Jihad: Understanding Cooperation among Terrorist Actors (New York: Columbia University Press, 2017); and Tricia Bacon, Why Terrorist Groups Form International Alliances (Philadelphia: University of Pennsylvania Press, 2018). 20. Rondeaux and Sterman, “Twenty-First Century Proxy Warfare,” p. 20. For examples of Cold War scholarship, see Karl W. Deutsch, “External Involvement in Internal War,” in Harry Eckstein, ed., Internal War: Problems and Approaches (New York: Free Press, 1964), pp. 100–110; Bertil Dunér, “Proxy Intervention in Civil Wars,” Journal of Peace Research, Vol. 18, No. 4 (December 1981), pp. 353–361, doi.org/10.1177%2F002234338101800404; Yaacov Bar-Siman-Tov, “The Strategy of War by Proxy,” Cooperation and Conflict, Vol. 19, No. 4 (November 1984), pp. 263–273, doi.org/10.1177%2F001083678401900405; and Naomi Joy Weinberger, Syrian Intervention in Lebanon (New York: Oxford University Press, 1986). 21. Daniel Byman and Sarah E. Kreps, “Agents of Destruction? Applying Principal-Agent Analysis to State-Sponsored Terrorism,” International Studies Perspectives, Vol. 11, No. 1 (February 2010), pp. 3–4, doi.org/10.1111/j.1528-3585.2009.00389.x; Idean Salehyan, Kristian Skrede Gleditsch, and David E. Cunningham, “Explaining External Support for Insurgent Groups,” International Organization, Vol. 65, No. 4 (Fall 2011), pp. 713–714, doi.org/10.1017/S0020818311000233; and Ariel I. Ahram, Proxy Warriors: The Rise and Fall of State-Sponsored Militias (Stanford, Calif.: Stanford University Press, 2011), p. 14. 22. Hughes, My Enemy's Enemy, pp. 22–23, 25; Mumford, Proxy Warfare, pp. 41–42; and Jeremy M. Berkowitz, “Delegating Terror: Principal-Agent Based Decision Making in State Sponsorship of Terrorism,” International Interactions, Vol. 44, No. 4 (2018), p. 715, doi.org/10.1080/03050629.2017.1414811; Salehyan, Gleditsch, and Cunningham, “Explaining External Support for Insurgent Groups,” p. 713; and Groh, Proxy War, p. 45. 23. Sabine C. Carey, Michael P. Colaresi, and Neil J. Mitchell, “Governments, Informal Links to Militias, and Accountability,” Journal of Conflict Resolution, Vol. 59, No. 5 (August 2015), pp. 852–853, doi.org/10.1177/0022002715576747; Byman and Kreps, “Agents of Destruction,” p. 6; and Borghard, “Friends with Benefits?” p. 18. As Daniel Byman and others point out, plausible deniability—often little more than a “convenient fiction”—works both ways and may allow a state attacked by a proxy to forgo retaliation against the latter's sponsor if it prefers to do so. See Byman, “Why Engage in Proxy War? A State's Perspective,” Lawfare blog, May 21, 2018, https://www.lawfareblog.com/why-engage-proxy-war-states-perspective. Furthermore, states do not always seek to conceal their involvement, but instead maintain proxy relationships as an “open secret.” See also Austin Carson, Secret Wars: Covert Conflict in International Politics (Princeton, N.J.: Princeton University Press), pp. 48–49. 24. Ahram, Proxy Warriors, p. 14. 25. Afshon Ostovar, “The Grand Strategy of Militant Clients: Iran's Way of War,” Security Studies, Vol. 28, No. 1 (2019), pp. 159–188, doi.org/10.1080/09636412.2018.1508862; and Ahram, Proxy Warriors. 26. Salehyan, “The Delegation of War to Rebel Organizations,” pp. 504–505; Byman and Kreps, “Agents of Destruction,” pp. 6–9; Salehyan, Gleditsch, and Cunningham, “Explaining External Support for Insurgent Groups,” pp. 714–715; Ahram, Proxy Warriors, pp. 14–15; Navin A. Bapat, “Understanding State Sponsorship of Militant Groups,” British Journal of Political Science, Vol. 42, No. 1 (December 2011), p. 4, doi.org/10.1017/S000712341100007X; Hughes, My Enemy's Enemy, pp. 49–51; Milos Popovic, “Fragile Proxies: Explaining Rebel Defection against Their State Sponsors,” Terrorism & Political Violence, Vol. 29, No. 5 (2017), pp. 924–925, doi.org/10.1080/09546553.2015.1092437; and Berkowitz, “Delegating Terror,” pp. 717–721. 27. We are grateful to an anonymous reviewer for this point. On covert signaling, see Austin Carson and Keren Yarhi-Milo, “Covert Communication: The Intelligibility and Credibility of Signaling in Secret,” Security Studies, Vol. 26, No. 1 (2017), pp. 124–156, org/10.1080/09636412.2017.1243921. 28. We thank the anonymous reviewers for this point. 29. Scholars of international relations have traditionally ascribed agency to states. In recent years, however, a growing body of scholarship has also applied agency to nonstate actors. See Bas Arts, “Nonstate Actors in Global Governance: Three Faces of Power,” working paper no. 2003/4 (Bonn, Germany: Max Planck Institute for Research on Collective Goods, April 2003); Kate O'Neill, Jörg Balsiger, and Stacy D. VanDeveer, “Actors, Norms, and Impact: Recent International Cooperation Theory and the Influence of the Agent-Structure Debate,” Annual Review of Political Science, Vol. 7 (June 2004), pp. 149–175, doi.org/10.1146/annurev.polisci.7.090803.161821; Deborah D. Avant, Martha Finnemore, and Susan K. Sell, eds., Who Governs the Globe? (New York: Cambridge University Press, 2010); and Naghmeh Nasiritousi, Mattias Hjerpe, and Björn-Ola Linnér, “The Roles of Nonstate Actors in Climate Change Governance: Understanding Agency through Governance Profiles,” International Environmental Agreements: Politics, Law, and Economics, Vol. 16, No. 1 (February 2016), pp. 109–126, org/10.1007%2Fs10784-014-9243-8. 30. Mumford, Proxy Warfare, pp. 45, 57. 31. Phillips and Valbj⊘rn, “‘What Is in a Name?'” 32. Rondeaux and Sterman, “Twenty-First Century Proxy Warfare,” pp. 50–51. 33. Carl von Clausewitz, On War, trans. and ed. Michael Howard and Peter Paret (Princeton, N.J.: Princeton University Press, 1984), p. 87. 34. For our present purposes, we adopt Sukanya Podder's definition of a legitimate nonstate actor as “the rightful wielder of power, maker and interpreter of rules or user of force and who thereby warrants support and compliance.” See Podder, “Understanding the Legitimacy of Armed Groups: A Relational Perspective,” Small Wars & Insurgencies, Vol. 28, Nos. 4–5 (September 2017), p. 687, doi.org/10.1080/09592318.2017.1322333. See also Mark C. Suchman, “Managing Legitimacy: Strategic and Institutional Approaches,” Academy of Management Review, Vol. 20, No. 3 (July 1995), pp. 571–610, doi.org/10.5465/amr.1995.9508080331. On the importance of legitimacy for nonstate actors, see Klaus Schlichte and Ulrich Schneckener, “Armed Groups and the Politics of Legitimacy,” Civil Wars, Vol. 17, No. 4 (2015), pp. 409–424, doi.org/10.1080/13698249.2015.1115573; and Isabelle Duyvesteyn, “Rebels and Legitimacy: An Introduction,” Small Wars & Insurgencies, Vol. 28, Nos. 4–5 (September 2017), pp. 669–685, doi.org/10.1080/09592318.2017.1322337. On the centrality of legitimacy for consolidating power in general, see David Beetham, The Legitimation of Power, 2nd ed. (New York: Palgrave Macmillan, 2013). On the critical importance of legitimacy for armed groups that seek to maintain power in the post-conflict phase, see, for example, Jeroen de Zeeuw, ed., From Soldiers to Politicians: Transforming Rebel Movements after Civil War (Boulder, Colo.: Lynne Rienner, 2008); and Veronique Dudouet, “From War to Politics: Resistance/Liberation Movements in Transition,” Berghof Report No. 17 (Berlin: Berghof Research Center for Constructive Conflict Management, April 2009). Furthermore, legitimacy is sought both domestically, with an eye toward deepening the group's local support base, and internationally, to increase the group's legitimacy vis-à-vis international actors. See, for example, Hyeran Jo, Compliant Rebels: Rebel Groups and International Law in World Politics (New York: Cambridge University Press, 2015); Stefan Malthaner, “Violence, Legitimacy, and Control: The Microdynamics of Support Relationships between Militant Groups and Their Social Environment,” Civil Wars, Vol. 17, No. 4 (2015), pp. 425–445, doi.org/10.1080/13698249.2015.1115575; and Tanisha M. Fazal and Margarita Konaev, “Homelands versus Minelands: Why Do Armed Groups Commit to the Laws of War?” Journal of Global Security Studies, Vol. 4, No. 2 (April 2019), pp. 149–168, org/10.1093/jogss/ogz005. 35. See Max Weber, “The Spirit of Work and Vocation-Second Lecture: Politics as Vocation, 1919,” in Tony Waters and Dagmar Waters, eds. and trans., Weber's Rationalism and Modern Society: New Translations on Politics, Bureaucracy, and Social Stratification (New York: Palgrave Macmillan, 2015), p. 136; Phil Williams, “Violent Nonstate Actors and National and International Security” (Zürich: International Relations and Security Networks, 2008); and Podder, “Understanding the Legitimacy of Armed Groups.” 36. As Groh and others have argued, such goals do not necessarily include the military defeat of a sponsor's adversary and instead might be limited to maintaining a status quo or creating chaos aimed at disrupting a status quo. See Groh, Proxy War, pp. 34–37. 37. Exceptions include Hezbollah, which is widely considered to be more militarily powerful than most states. As discussed later, we show that our argument applies even in the “hard case” of Hezbollah, albeit with some qualification. 38. Williams, “Violent Nonstate Actors and National and International Security,” p. 13. 39. Barak Mendelsohn, Jihadism Constrained: The Limits of Transnational Jihadism and What It Means for Counterterrorism (Lanham, Md.: Rowman and Littlefield, 2018). 40. Thomas Risse, ed., Governance without a State? Policies and Politics in Areas of Limited Statehood (New York: Columbia University Press, 2011); Ulrich Schneckener, “Spoilers or Governance Actors? Engaging Armed Nonstate Groups in Areas of Limited Statehood,” SFB-Governance Working Paper Series No. 21 (Berlin: DFG Research Center [SFB] 700, October 2009); and Keith Krause and Jennifer Milliken, “Introduction: The Challenge of Nonstate Armed Groups,” Contemporary Security Policy, Vol. 30, No. 2 (2009), pp. 202–220, org/10.1080/13523260903077296. 41. Williams, “Violent Nonstate Actors and National and International Security,” p. 12. 42. Groh, Proxy War, p. 26. 43. Scott Sigmund Gartner and Gary M. Segura, “War, Casualties, and Public Opinion,” Journal of Conflict Resolution, Vol. 42, No. 3 (June 1998), pp. 278–300, org/10.1177/0022002798042003004. 44. The United Nations (UN) Charter states in Article 2.4 that “all members shall refrain in their international relations from the threat or use of force against the territorial integrity or political independence of any state, or in any other manner inconsistent with the Purposes of the United Nations.” See UN, Charter of the United Nations, 1 UNTS XVI, October 24, 1945, https://www.un.org/en/sections/un-charter/un-charter-full-text/. 45. Kelly M. Kadera, “Transmission, Barriers, and Constraints: A Dynamic Model of the Spread of War,” Journal of Conflict Resolution, Vol. 42, No. 3 (June 1998), pp. 367–368, org/10.1177/002200 2798042003008. 46. On the costs of Hezbollah's deployment to Syria, see, for example, Colin P. Clarke, “A Glass Half Empty? Taking Stock of Hezbollah's Losses in Syria,” Jerusalem Post, October 15, 2017, https://www.jpost.com/Opinion/A-glass-half-empty-Taking-stock-of-Hezbollahs-losses-in-Syria-507497. This is not to suggest that communities supporting these nonstate actors are immune to the impact of casualty rates, only that nonstate actors will be less responsive to such complaints, and that such factors will have less impact on their decision to employ proxies. 47. Ami Pedahzur, Suicide Terrorism (Cambridge: Polity, 2005); Assaf Moghadam, The Globalization of Martyrdom: Al Qaeda, Salafi Jihad, and the Diffusion of Suicide Attacks (Baltimore, Md.: Johns Hopkins University Press, 2008); and Jeffrey William Lewis, The Business of Martyrdom: A History of Suicide Bombing (Annapolis: Naval Institute Press, 2012). On the motivations of foreign fighters, see David Malet, Foreign Fighters: Transnational Identity in Civil Conflicts (New York: Oxford University Press, 2013); and Daniel Byman, Road Warriors: Foreign Fighters in the Armies of Jihad (New York: Oxford University Press, 2019). 48. Alan J. Kuperman, “Rethinking the Responsibility to Protect,” Whitehead Journal of Diplomacy and International Relations, Vol. 10, No. 1 (Winter/Spring 2009), pp. 22–24, https://heinonline.org/HOL/P?h=hein.journals/whith10&i=33. 49. On the provocation strategy, see Carlos Marighella, Minimanual of the Urban Guerrilla (Montreal and Toronto: Abraham Guillen Press and Arm the Spirit, 2002); Martha Crenshaw, “The Causes of Terrorism,” Comparative Politics, Vol. 13, No. 4 (July 1981), pp. 379–399, doi.org/10.2307/421717; Andrew H. Kydd and Barbara F. Walter, “The Strategies of Terrorism,” International Security, Vol. 31, No. 1 (Summer 2006), pp. 49–80, doi.org/10.1162/isec.2006.31.1.49; and Peter R. Neumann and M.L.R. Smith, The Strategy of Terrorism: How It Works and Why It Fails (London: Routledge, 2008). 50. Jeremy Weinstein, Inside Rebellion: The Politics of Insurgent Violence (New York: Cambridge University Press, 2007). 51. Mao Tse-Tung, On Guerrilla Warfare, trans. Samuel B. Griffith II (Champaign: University of Illinois Press, 2000); Sir Robert Thompson, Defeating Communist Insurgency: Experiences from Malaya and Vietnam (New York: Frederick A. Praeger, 1966); David Galula, Counterinsurgency Warfare: Theory and Practice (Westport, Conn.: Praeger Security International, 2006); Ernesto Che Guevara, Guerilla Warfare (Lincoln: Bison, 1998); David H. Petraeus, James F. Amos, and John A. Nagl, The U.S. Army/Marine Corps Counterinsurgency Field Manual (Chicago: University of Chicago Press, 2007); Bard E. O'Neill, Insurgency and Terrorism: Inside Modern Revolutionary Warfare (McLean, Va.: Brassey's, 1990); David Kilcullen, The Accidental Guerrilla: Fighting Small Wars in the Midst of a Big One (New York: Oxford University Press, 2009); and Norma J. Kriger, Zimbabwe's Guerrilla War: Peasant Voices (Cambridge: Cambridge University Press, 1991). 52. Assaf Moghadam and Brian Fishman, Fault Lines in Global Jihad: Organizational, Strategic, and Ideological Fissures (London: Routledge, 2011); Schlichte and Schneckener, “Armed Groups and the Politics of Legitimacy,” pp. 416–419; and Ana Arjona, “Civilian Cooperation and Non-Cooperation with Nonstate Armed Groups: The Centrality of Obedience and Resistance,” Small Wars & Insurgencies, Vol. 28, Nos. 4–5 (September 2017), pp. 760–761, org/10.1080/09592318.2017.1322328. 53. For example, nonstate actors seek to provide social services to a population to help them enhance their legitimacy. Alexus G. Grynkewich, “Welfare as Warfare: How Violent Nonstate Groups Use Social Services to Attack the State,” Studies in Conflict & Terrorism, Vol. 31, No. 4 (April 2008), pp. 350–370, org/10.1080/10576100801931321. 54. In fact, their external support may even diminish some of their proxies’ domestic legitimacy in the eyes of the local population, which perceives them as agents of a foreign power. See Salehyan, “The Delegation of War to Rebel Organizations,” p. 507. 55. Hezbollah is believed to have had a fighting force of 45,000 in 2016, including 25,000 reservists and part-time fighters, of which between 7,000 and 10,000 are believed to have been deployed in Syria. It reportedly has at its disposal over 100,000 rockets and hundreds of drones, as well as advanced weapons systems such as surface-to-sea, anti-tank, and anti-air missiles, including SA-22 systems. See Amos Harel and Gili Cohen, “Hezbollah: From Terror Group to Army,” Haaretz, July 12, 2016, https://www.haaretz.com/st/c/prod/eng/2016/07/lebanon2/; and Seth G. Jones and Maxwell B. Marcusen, “The Escalating Conflict with Hezbollah in Syria,” CSIS Brief (Washington, D.C.: Center for Strategic and International Studies, June 2018), p. 3. Hezbollah is frequently described as the most capable armed non-state actor in the world. See, for example, Shaan Shaikh and Ian Williams, “Hezbollah's Missiles and Rockets,” CSIS Brief (Washington, D.C.: Center for Strategic and International Studies, July 2018), p. 1; and Nicholas Blanford, “Hezbollah: In Syria for the Long Haul” (Washington, D.C.: Middle East Institute, November 18, 2014), https://www.mei.edu/publications/hezbollah-syria-long-haul. As far as AQAP is concerned, the numbers of its fighters likely never exceeded 4,000. See Elisabeth Kendall, “Contemporary Jihadi Militancy in Yemen: How Is the Threat Evolving?” MEI Policy Paper No. 2018-7 (Washington, D.C.: Middle East Institute, July 2018), p. 5; and Bureau of Counterterrorism, “Country Reports on Terrorism, 2017“ (Washington, D.C.: United States Department of State, September 2018), p. 328. Unlike Hezbollah and the Syrian Democratic Forces, AQAP does not receive state support. The size of the YPG's forces was estimated at 60,000 fighters at the end of 2016, but the SDF is not nearly as well equipped militarily as Hezbollah. See Tom Perry, “Exclusive: Syrian Kurdish YPG Aims to Expand Force to Over 100,000,” Reuters, March 20, 2017, https://www.reuters.com/article/us-mideast-crisis-syria-ypg-exclusive/exclusive-syrian-kurdish-ypg-aims-to-expand-force-to-over-100000-id USKBN16R1QS; and Idrees Ali, Lesley Wroughton, and Jonathan Landay, “Exclusive: U.S. Commanders Recommend Letting Kurdish Fighters in Syria Keep Weapons,” Reuters, December 29, 2018, https://www.reuters.com/article/us-mideast-crisis-syria-usa-exclusive/exclusive-u-s-commanders-recommend-letting-kurdish-fighters-in-syria-keep-weapons-idUSKCN1OR1OD. 56. See Nicholas Blanford, “Israel: Hezbollah Is Now Stronger Than Any Arab Army,” Christian Science Monitor, June 9, 2014, https://www.csmonitor.com/World/Middle-East/2014/0609/Israel-Hezbollah-is-now-stronger-than-any-Arab-army. 57. Although Hezbollah does not promulgate its budget, most recent estimates indicate that Iranian support constitutes around 70 percent to 80 percent of it, which would amount to approximately700 million. See Ali Bakeer, “Hezbollah's Finances Are Its Achilles’ Heel,” National Interest, January 27, 2019, https://nationalinterest.org/feature/hezbollahs-finances-are-its-achilles-heel-42462. This figure does not include an estimated $40 million a month that Hezbollah draws from its own global sources of revenue. See Nicholas Blanford, “Hezbollah's Evolution: From Lebanese Militia to Regional Player,” MEI Policy Paper 2017–4 (Washington, D.C.: Middle East Institute, November 2017), p. 3. The budget of the YPG is equally difficult to ascertain. According to an analysis by the Omran Institute for Strategic Studies, in 2014 the Democratic Autonomous Administration had a total revenue of 3.7 billion Syrian pounds (about$5.8 million) of which 50 percent went to “self-defense and protection.” See Bedir Mulla Rashid, “The Autonomous Administration in Northern Syria: Questions of Legitimacy and Identity” (Istanbul: Omran for Strategic Studies, July 2018), p. 11. In 2019, the Pentagon requested $300 million for the Counter-Islamic State in Iraq and Syria Train and Equip Fund. See Ercan Gurses and Daren Butler, “U.S. Funding of Syrian YPG Militia Will Impact Turkey's Decisions: Erdogan,” Reuters, February 13, 2018, https://www.reuters.com/article/us-mideast-crisis-syria-turkey-usa/u-s-funding-of-syrian-ypg-militia-will-impact-turkeys-decisions-erdogan-idUSKBN1FX0XP; and Laurie Mylroie, “U.S. Clarifies Confusion over 2019 Funds for SDF,” Kurdistan 24, February 14, 2018, https://www.kurdistan24.net/en/news/c4698dc4-f618-4099-b50f-de6a6c38cc40. AQAP's funding has historically derived mostly from theft, oil and gas revenue, kidnapping operations, and donations. In April 2015, the group reportedly stole 13 billion rials ($60 million) and $1.5 million from the Central Bank in al-Mukallah. Bureau of Counterterrorism, “Country Reports on Terrorism, 2017” (Washington D.C.: U.S. Department of State, September 2018), p. 328. According to the UN Security Council, the group also had access to an estimated$2 million per day in revenue from taxes from shippers and traders while it occupied al-Mukallah from April 2015 to April 2016. UN Security Council, “Eighteenth Report of the Analytical Support and Sanctions Monitoring Team Submitted Pursuant to Resolution 2253 (2015) Concerning Islamic State in Iraq and the Levant (Da'esh), Al-Qaida, and Associated Individuals and Entities,” July 19, 2016, pp. 10–11, https://www.un.org/en/ga/search/view_doc.asp?symbol=S/2016/629.

58.

Thirteen of these seats went to Hezbollah itself, making it the second strongest party after the Free Patriotic Movement, with which it cooperates. “Factbox: Hezbollah and Allies Gain Sway in Lebanon Parliament,” Reuters, May 22, 2018, https://www.reuters.com/article/us-lebanon-election-parliament-factbox/factbox-hezbollah-and-allies-gain-sway-in-lebanon-parliament-idUSKCN1IN1OJ. In addition, a number of governments, including Germany's, have acknowledged the key role Hezbollah plays within Lebanese politics and subsequently refused calls to outlaw Hezbollah as a whole, instead calling for “dialogue” and banning only its “military wing.” “Niels Annen über Hisbollah: ‘Wir setzen auf Dialog'” [Niels Annen on Hezbollah: “We count on dialogue”], Spiegel Online, March 8, 2019.

59.

The YPG has garnered some limited international legitimacy, particularly as it proved itself as a valuable fighting force against the Islamic State, but its international backers have refrained from publicly endorsing Kurdish autonomy in northern Syria. AQAP enjoys some pockets of support in the areas under its control, but little legitimacy in the international arena. As far as AQAP's support in Yemen is concerned, the International Crisis Group (ICG) reported that “residents in AQAP-controlled areas, while not supporting the group's ideology, have regularly praised its prioritization of security, basic services, and a mechanism to resolve grievances, such as long-running land disputes.” Elsewhere, the report states that “AQAP is an internally diverse organisation with varying layers of support among the local population.” See ICG, “Yemen's al-Qaida: Expanding the Base,” Middle East Report No. 174 (Brussels: ICG, February 2017), pp. 11, 26.

60.

On crucial case designs, see Harry Eckstein, “Case Studies and Theory in Political Science,” in Fred Greenstein and Nelson Polsby, eds., Handbook of Political Science, vol. 7 (Reading, Mass.: Addison-Wesley, 1975), pp. 79–138; and Jack S. Levy, “Case Studies: Types, Designs, and Logics of Inference,” Conflict Management and Peace Science, Vol. 25, No. 1 (2008), pp. 1–18, org/10.1080/07388940701860318.

61.

Alexander L. George and Andrew Bennett, Case Studies and Theory Development in the Social Sciences (Cambridge, Mass.: MIT Press, 2005), pp. 61–63.

62.

ICG, “Yemen's al-Qaida,” pp. 3–4; and Kendall, “Contemporary Jihadi Militancy in Yemen,” p. 3.

63.

ICG, “Yemen's al-Qaida,” p. 4.

64.

Gabriel Koehler-Derrick, “The Abbottabad Documents: Bin Ladin's Cautious Strategy in Yemen,” CTC Sentinel, Vol. 5, No. 5 (May 2012), pp. 15–19.

65.

66.

Michael Horton, “Fighting the Long War: The Evolution of al-Qa'ida in the Arabian Peninsula,” CTC Sentinel, Vol. 10, No. 1 (January 2017), pp. 17–22.

67.

ICG, “Yemen's al-Qaida,” p. 6.

68.

Nadwa al-Dawsari, “Our Common Enemy: Ambiguous Ties between al-Qaida and Yemen's Tribes” (Beirut: Carnegie Middle East Center, January 11, 2018), https://carnegie-mec.org/2018/01/11/our-common-enemy-ambiguous-ties-between-al-Qaida-and-yemen-s-tribes-pub-75225; and Tik Root, “Al-Qaida Destroyed Our Family,” Slate, February 28, 2014, https://slate.com/news-and-politics/2014/02/al-qaida-destroyed-the-al-dhahab-family-how-a-yemeni-family-lost-four-sons-to-the-fight.html.

69.

Horton, “Fighting the Long War,” p. 19; and al-Dawsari, “Our Common Enemy.”

70.

Horton, “Fighting the Long War,” p. 19.

71.

Saeed Al Batati, “Yemen: The Truth behind al-Qaida's Takeover of Mukalla,” Al Jazeera, September 16, 2015, https://www.aljazeera.com/news/2015/09/yemen-truth-al-Qaida-takeover-mukalla-150914101527567.html.

72.

Ibid.; and Faisal Edroos and Saleh Al Batati, “After al-Qaida: No Signs of Recovery in Yemen's Mukalla,” al-Jazeera, January 11, 2018, https://www.aljazeera.com/indepth/features/al-Qaida-signs-recovery-yemen-mukalla-180111135554851.html.

73.

Al Batati, “Yemen: The Truth behind al-Qaida's Takeover of Mukalla”; and “MEE Exclusive: Al-Qaida Pulls Out of Key Government Buildings in Southeastern Yemen,” Middle East Eye, May 12, 2015, https://www.middleeasteye.net/news/mee-exclusive-al-Qaida-pulls-out-key-government-buildings-southeastern-yemen.

74.

Studies typically estimate that AQAP's number of core fighters never exceeded 4,000. See Kendall, “Contemporary Jihadi Militancy in Yemen,” p. 5. When it captured al-Mukalla, the group seized large quantities of military equipment, including tanks and rocket launchers. See Horton, “Fighting the Long War,” p. 21; and Associated Press, “Al Qaeda in Yemen Seizes Huge Weapons Depot from Army,” CBS News, April 17, 2015, https://www.cbsnews.com/news/al-qaeda-in-yemen-seizes-huge-weapons-depot-from-army/.

75.

Al Batati, “Yemen: The Truth behind al-Qaida's Takeover of Mukalla.”

76.

Ibid.; and Associated Press, “MEE EXCLUSIVE: Al-Qaida Pulls Out of Key Government Buildings in Southeastern Yemen.”

77.

“MEE Exclusive: Al-Qaida Pulls Out of Key Government Buildings in Southeastern Yemen.”

78.

Ibid.

79.

Horton, “Fighting the Long War,” p. 19.

80.

Ibid., pp. 19–21.

81.

Al Batati, “Yemen: The Truth behind al-Qaida's Takeover of Mukalla.”

82.

ICG, “Yemen's al-Qaida,” p. 14.

83.

Elisabeth Kendall, “Al-Qaida and Islamic State in Yemen,” in Simon Staffell and Akil Awan, eds., Jihadism Transformed: Al-Qaida and Islamic State's Global Battle of Ideas (New York: Oxford University Press, 2016), pp. 91–92.

84.

Michael Horton, “Guns for Hire: How al-Qaeda in the Arabian Peninsula Is Securing Its Future in Yemen,” Jamestown Terrorism Monitor, Vol. 16, No. 2 (January 2018), p. 9.

85.

Edroos and Al Batati, “After al-Qaida.”

86.

Horton, “Guns for Hire,” p. 9.

87.

Nadwa Al-Dawsari, “Foe Not Friend: Yemeni Tribes and Al-Qaeda in the Arabian Peninsula,” (Washington, D.C.: Project on Middle East Democracy, February 2018), p. 22.

88.

Jim Garamone, “Coalition Trains Kurdish Troops for Anti-ISIL Fight,” U.S. Department of Defense, October 20, 2015, https://www.defense.gov/News/Article/Article/624865/coalition-trains-kurdish-troops-for-anti-isil-fight/; Constanze Letsch, “U.S. Drops Weapons and Ammunition to Help Kurdish Fighters in Kobani,” Guardian, October 20, 2014, https://www.theguardian.com/world/2014/oct/20/turkey-iraqi-kurds-kobani-isis-fighters-us-air-drops-arms; and Michael Stephens and Aaron Stein, “The YPG: America's New Best Friend,” Al Jazeera, June 28, 2015, https://www.aljazeera.com/indepth/opinion/2015/06/ypg-america-friend-isil-kurds-syria-150627073034776.html.

89.

Steven A. Cook, “Who Exactly Are ‘the Kurds'?” Atlantic, February 25, 2016, https://www.theatlantic.com/international/archive/2016/02/kurds-turkey-pkk-ypg/470991/; Kadir Ustun, “U.S. Alliance with Syrian PYD Alienates Turkey,” Al Jazeera, June 2, 2016, https://www.aljazeera.com/indepth/opinion/2016/06/alliance-syrian-pyd-alienates-turkey-160601095726203.html; and Aaron Stein, “Reconciling U.S.-Turkish Interests in Northern Syria” (New York: Council on Foreign Relations, February 2017), pp. 3–5.

90.

“The Kurds’ Precarious Balancing Act in Syria,” Strategic Comment, Vol. 23, No. 4 (2017), pp. v–vi, doi.org/10.1080/13567888.2017.1334853; Vittoria Federici, “The Rise of Rojava: Kurdish Autonomy in the Syrian Conflict,” SAIS Review of International Affairs, Vol. 35, No. 2 (Summer–Fall 2015), p. 83, doi.org/10.1353/sais.2015.0023; and Michiel Leezenberg, “The Ambiguities of Democratic Autonomy: The Kurdish Movement in Turkey and Rojava,” Southeast European and Black Sea Studies, Vol. 16, No. 4 (October 2016), pp. 675–678, 683, doi.org/10.1080/14683857.2016.1246529. See also ICG, “Flight of Icarus? The PYD's Precarious Rise in Syria,” Middle East Report No. 151 (Brussels: ICG, May 2014).

91.

ICG, “The PKK's Fateful Choice in Northern Syria,” Middle East Report No. 176 (Brussels: ICG, May 2017), p. 12; and Aron Lund, “Origins of the Syrian Democratic Forces: A Primer,” Syria Deeply, January 22, 2016, https://www.newsdeeply.com/syria/articles/2016/01/22/origins-of-the-syrian-democratic-forces-a-primer.

92.

ICG, “The PKK's Fateful Choice in Northern Syria,” pp. 12–13.

93.

Lund, “Origins of the Syrian Democratic Forces.”

94.

Exact figures are not available, but most studies estimate that the YPG/YPJ have at least 25,000 members. Membership in the other groups is mostly in the low hundreds, with a few exceptions such as the Syriac Military Council and the Sanadid Forces. See, for example, Genevieve Casagrande, “The Road to ar-Raqqah: Background on the Syrian Democratic Forces” (Washington, D.C.: Institute for the Study of War, November 2016), pp. 1–2; and “Syrian Democratic Forces (SDF): From the Washington-Moscow Agreement to Animosity with Turkey” (Riyadh: King Faisal Center for Research and Islamic Studies, August–September 2016), p. 11.

95.

ICG, “The PKK's Fateful Choice in Northern Syria,” p. 13.

96.

Aymenn Jawad Al-Tamimi, “Christian Militia and Political Dynamics in Syria,” Syria Comment, February 23, 2018, https://www.joshualandis.com/blog/christian-militia-political-dynamics-syria/.

97.

ICG, “The PKK's Fateful Choice in Northern Syria,” p. 16.

98.

MFS's all-female unit is called the Bethnarin Women's Protection Forces. See also Pascal Andresen, “In This Sign They Will Conquer? Christian Militias in the Syrian Conflict,” Bellingcat, November 4, 2017, https://www.bellingcat.com/news/mena/2017/11/04/sign-will-conquer-christian-militias-syrian-conflict/.

99.

Wladimir van Wilgenburg, “Kurdish Strategy towards Ethnically-Mixed Areas in the Syrian Conflict,” Jamestown Terrorism Monitor, Vol. 11, No. 23 (December 2013), p. 7. Aymenn al-Tamimi, an expert on the Islamic State and on militant groups in Syria, independently confirmed that Jabhat al-Akrad was a proxy of YPG, whom the latter used mainly for outreach to some rebel factions in the North Aleppo countryside. Email correspondence with al-Tamimi, June 8, 2018. It is noteworthy that a Jabhat al-Akrad commander denied these claims in 2014, explaining that they had no partnership whatsoever with the YPG. “Interview with Jabhat al-Akrad Commander,” Rojava Report, April 20, 2014, https://rojavareport.wordpress.com/2014/04/20/interview-with-jabhat-al-akrad-commander/.

100.

Stein, “Reconciling U.S.-Turkish Interests in Northern Syria,” p. 7.

101.

ICG, “The PKK's Fateful Choice in Northern Syria,” pp. 13, 16.

102.

However, a small number of MFS fighters had been involved in the battle for Raqqa. See Andresen, “In This Sign They Will Conquer.”

103.

Ibid.

104.

“Syrian-Kurdish SDF Successfully Absorbing Non-Kurdish groups, Says U.S.,” Rudaw, March 9, 2016, https://www.rudaw.net/english/middleeast/syria/090320161.

105.

Alfred Hackensberger, “Dieser Scheich will den Islamischen Staat bezwingen” [This sheikh wants to overcome the Islamic State”], Welt, December 7, 2015, https://www.welt.de/politik/ausland/article149693867/Dieser-Scheich-will-den-Islamischen-Staat-bezwingen.html.

106.

Marwan Hisham, “I Fled the Islamic State's ‘Caliphate’ in Raqqa—But Fear Its Liberators,” Foreign Policy, June 23, 2017, https://foreignpolicy.com/2017/06/23/i-fled-the-islamic-states-caliphate-in-raqqa-but-fear-its-liberators/; and Andrew J. Tabler, “Eyeing Raqqa: A Tale of Four Tribes,” Policy Notes No. 39 (Washington, D.C.: Washington Institute for Near East Policy, 2017).

107.

Stein, “Reconciling U.S.-Turkish Interests in Northern Syria,” p. 7.

108.

Shervan Derwish, “The Safe Zone Northern Syria Needs,” New York Times, January 23, 2019, https://www.nytimes.com/2019/01/23/opinion/syria-turkey-trump.html.

109.

Daniel Wilkofsky and Khalid Fatah, “Northern Syria's Anti-Islamic State Coalition Has an Arab Problem,” War on the Rocks blog, September 18, 2017, https://warontherocks.com/2017/09/northern-syrias-anti-islamic-state-coalition-has-an-arab-problem/.

110.

As the report notes, “Official rhetoric signals inclusiveness and pluralism, but YPG flags and posters of Öcalan adorn streets and town squares (including in majority-Arab areas) in a manner typical of autocratic, single-party rule elsewhere in the region. Arab figures willing to participate in the self-administration are handed impressive titles but no real authority. Local governance bodies function as channels to convey complaints and petitions rather than as platforms for effective participation, while ultimate power of decision rests with the Qandil-trained PKK cadres.” See ICG, “Fighting ISIS: The Road To and Beyond Raqqa,” Middle East Briefing No. 53 (Brussels: ICG, April 2017), p. 4.

111.

“Manbij Military Council: We Can Defend Our City,” ANF News, June 6, 2018, https://anfenglish.com/news/manbij-military-council-we-can-defend-our-city-27250.

112.

Liz Sly, “U.S. Military Aid Is Fueling Big Ambitions for Syria's Leftist Kurdish Militia,” Washington Post, January 7, 2017, https://www.washingtonpost.com/world/middle_east/us-military-aid-is-fueling-big-ambitions-for-syrias-leftist-kurdish-militia/2017/01/07/6e457866-c79f-11e6-acda-59924caa2450_story.html.

113.

Wilkofsky and Fatah, “Northern Syria's Anti-Islamic State Coalition Has an Arab Problem.”

114.

“U.S.-Backed SDF Captures Syria's Largest Oil Field from ‘Islamic State,'” DW, October 23, 2017, https://www.dw.com/en/us-backed-sdf-captures-syrias-largest-oil-field-from-islamic-state/a-41066879.

115.

Casey L. Addis and Christopher M. Blanchard, “Hezbollah: Background and Issues for Congress,” CRS Report No. R41446 (Washington, D.C.: Congressional Research Service, January 2011), p. 19.

116.

Daniel Byman, Deadly Connections: States that Sponsor Terrorism (New York: Cambridge University Press, 2007), pp. 87–88.

117.

James Risen, “A Region Inflamed: The Hand of Tehran; Hezbollah, in Iraq, Refrains From Attacks on Americans,” New York Times, November 24, 2003, https://www.nytimes.com/2003/11/24/world/region-inflamed-hand-tehran-hezbollah-iraq-refrains-attacks-americans.html; Joseph Felter and Brian Fishman, “Iranian Strategy in Iraq: Politics and ‘Other Means'” (West Point, N.Y.: Combating Terrorism Center at West Point, October 2008), pp. 26–27, 56; Dexter Filkins, “The Shadow Commander,” New Yorker, September 30, 2013, https://www.newyorker.com/magazine/2013/09/30/the-shadow-commander; Alexander Corbeil and Amarnath Amarasingam, “Houthi Hezbollah: Iran's Train-and-Equip Program in Sanaa,” Foreign Affairs, March 31, 2016, https://www.foreignaffairs.com/articles/2016-03-31/houthi-hezbollah; Joshua Koontz, “Iran's Growing Casualty Count in Yemen,” War on the Rocks blog, June 1, 2017, https://warontherocks.com/2017/06/irans-growing-casualty-count-in-yemen/; Jeffrey White, “Hizb Allah at War in Syria: Forces, Operations, Effects, and Implications,” CTC Sentinel, Vol. 7, No. 1 (January 2014), pp. 14–18; Jones and Markusen, “The Escalating Conflict with Hezbollah in Syria,” pp. 2–3; Ali Soufan, “Qassem Soleimani and Iran's Unique Regional Strategy,” CTC Sentinel, Vol. 11, No. 10 (November 2018), pp. 1–12; and Ostovar, “The Grand Strategy of Militant Clients,” pp. 175–176.

118.

Matthew Levitt and Philip Smyth, “Kataib al-Imam Ali: Portrait of an Iraqi Shiite Militant Group Fighting ISIS” (Washington, D.C.: Washington Institute for Near East Policy, January, 2015); David Daoud, “Hezbollah Fighters Train Iraqi Shiite Militants near Mosul,” Long War Journal, November 5, 2016, https://www.longwarjournal.org/archives/2016/11/hezbollah-fighters-train-iraqi-shiite-militants-near-mosul.php; and Garrett Nada and Mattisan Rowan, “Part 2: Pro-Iran Militias in Iraq” (Washington, D.C.: Iran Primer, Wilson Center, April 26, 2018), https://iranprimer.usip.org/blog/2018/apr/26/part-2-pro-iran-militias-iraq.

119.

Mona Alami, “Hezbollah Embedded in Syria” (Washington, D.C.: Atlantic Council, March 2, 2017), https://www.atlanticcouncil.org/blogs/syriasource/hezbollah-is-embedded-in-syria/; Daoud, “Hezbollah Fighters Train Iraqi Shiite Militants near Mosul;” and Karen DeYoung and Joby Warrick, “Iran, Hezbollah Build Militia Networks in Syria in Event that Assad Falls, Officials Say,” Washington Post, February 10, 2013, https://www.washingtonpost.com/world/national-security/iran-hezbollah-build-militia-networks-in-syria-in-event-that-assad-falls-officialssay/2013/02/10/257a41c8-720a-11e2-ac36-3d8d9dcaa2e2_story.html.

120.

In the past, Hezbollah also provided training to al-Qaida fighters in its camps in the Beqaa Valley, yet the argument that al-Qaida thus became a proxy of Hezbollah is a stretch. See Cathy Scott-Clark and Adrian Levy, The Exile: The Flight of Osama Bin Laden (London: Bloomsbury 2017), p. 88.

121.

E. al-Rihani, “Long Live ‘Saraya,'” Al-Manar, March 15, 2011, http://archive.almanar.com.lb/english/article.php?id=6203. The exact date of Nasrallah's announcement is disputed. See Nicholas Blanford, God's Warriors: Inside Hezbollah's Thirty Year Struggle against Israel (New York: Random House, 2011), p. 197.

122.

Blanford, God's Warriors, p. 197.

123.

Ibid.

124.

Al-Rihani, “Long Live ‘Saraya.'”

125.

Blanford, God's Warriors, pp. 212–214.

126.

Ibid., p. 439.

127.

Alexandra Masi, “Christian, Sunni, and Shia: Meet Hezbollah's Non-Denominational Military Branch Defending Lebanon, Fighting in Syria,” International Business Times, November 4, 2015, https://www.ibtimes.com/christian-sunni-shia-meet-hezbollahs-non-denominational-military-branch-defending-2169257.

128.

“Lebanon's Border with Syria: Taking Charge,” Economist, June 15, 2014, https://www.economist.com/pomegranate/2014/06/15/taking-charge.

129.

Ahmad Nizar Hamzeh, In the Path of Hizballah (Syracuse, N.Y.: Syracuse University Press, 2004), p. 91. According to the Hezbollah outlet al-Manar, the number of attacks carried out by Saraya was closer to 400. See al-Rihani, “Long Live ‘Saraya.'”

130.

Blanford, God's Warriors, p. 213.

131.

Ibid.

132.

Chris Zambelis, “Hizb Allah's Lebanese Resistance Brigades,” CTC Sentinel, Vol. 7, No. 11 (November–December 2014), p. 10. Some scholars dispute that Hezbollah has become solely committed to Lebanese interests. See Robert G. Rabil, “Hezbollah, the Islamic Association, and Lebanon's Confessional System: Al-Infitah and Lebanonization,” Levantine Review, Vol. 1, No. 1 (May 2012), pp. 49–67, https://doi.org/10.6017/lev.v1i1.2151; and Eitan Azani, “Hezbollah's Strategy of ‘Walking on the Edge': Between Political Game and Political Violence,” Studies in Conflict & Terrorism, Vol. 35, No. 11 (October 2012), pp. 741–759, org/10.1080/1057610X.2012.720238.

133.

Blanford, God's Warriors, p. 439.

134.

Ibid., p. 440.

135.

Nadav Pollak, “The Transformation of Hezbollah by Its Involvement in Syria,” Research Notes No. 35 (Washington, D.C.: Washington Institute for Near East Policy, August 2016), p. 4.

136.

Masi, “Christian, Sunni, and Shia.”

137.

Ibid.

138.

“Lebanon's Border with Syria;” and Masi, “Christian, Sunn, and Shia.”

139.

Masi, “Christian, Sunni, and Shia;” and Pollak, “The Transformation of Hezbollah by Its Involvement in Syria,” p. 4.

140.

Nizar Hamzeh, In the Path of Hizballah, p. 78.

141.

Masi, “Christian, Sunni, and Shia.”

142.

See, for example, Joseph Elie Alagha, The Shifts in Hizballah's Ideology: Religious Ideology, Political Ideology, and Political Program (Leiden, Netherlands: Amsterdam University Press, 2006), pp. 149–189; Amal Saad-Ghorayeb, Hizbu'llah: Politics and Religion (London: Pluto, 2002), pp. 78–87; and Nizar Hamzeh, In the Path of Hizballah, pp. 108–135.

143.

Zambelis, “Hizb Allah's Lebanese Resistance Brigades,” p. 11.

144.

Although some Arab units took part in frontline operations, their importance remains negligible. This was true even in the battle for Raqqa. See John Davison, “Arab Fighters Struggle to Assert Role in Raqqa Assault,” Reuters, August 16, 2017, https://www.reuters.com/article/us-mideast-crisis-syria-raqqa/arab-fighters-struggle-to-assert-role-in-raqqa-assault-idUSKCN1AW1L4.

145.

ICG, “Yemen's al-Qaida,” p. 14.

146.

ICG, “Fighting ISIS,” p. 4.

147.

An additional reason for Hezbollah's immunity to defection is that the group plays according to the rules of the political game in Lebanon. Unlike AQAP and YPG, it does not try to carve out and govern territory. Although it provides services in areas dominated by Shia, it does so within the framework of the Lebanese state. See, for example, Benedetta Berti, “Lebanon,” in Assaf Moghadam, ed., Militancy and Political Violence in Shiism: Trends and Patterns (Oxford: Routledge, 2012).

148.

Hughes, My Enemy's Enemy, p. 12; Loveman, “Assessing the Phenomenon of Proxy Intervention,” p. 32; Mumford, Proxy Warfare, pp. 61–66; and Salehyan, “The Delegation of War to Rebel Organizations,” p. 507.

149.

This is not to suggest that state sponsorship will disappear, or even decline. The different forms of sponsorship are likely to coexist in the future.