## Abstract

In the midst of civil war, rebel groups often expend significant resources opening offices in foreign capitals, meeting with heads of state, expanding their overseas networks, appealing to international organizations, and contacting foreign media. Existing scholarship has generally neglected international diplomacy as an aspect of violent rebellion, focusing instead on rebel efforts at domestic organization. A systematic documentation of rebel diplomacy in post–1950 civil wars using new quantitative and qualitative data shows that rebel diplomacy is commonplace and that many groups demonstrate as much concern for overseas political campaigns as they do for domestic and local mobilization. Diplomacy, furthermore, is not a weapon of the militarily weak, but a tactical choice for rebel groups seeking political capital within an international system that places formidable barriers to entry on nonstate entities. An original analysis of the diplomacy of the National Union for the Total Independence of Angola in the Angolan civil war using archival sources further demonstrates why rebels may become active diplomats in one phase of a conflict but eschew diplomacy in another. More broadly, the international relations of rebel groups promise to be an important new research agenda in understanding violent politics.

## Introduction

Violent rebels are often active diplomats. While opposition forces fought against the Syrian regime of Bashar al-Assad, the head of the Syrian National Coalition, Ahmad al-Jarba, made his first official American tour in May 2014.1 During his visit, the coalition achieved a major feat when the State Department recognized the rebels' Washington office as a foreign mission, thus conferring the group diplomatic status in the United States.2 Such events are far from unique in civil wars. While fighting a notoriously brutal civil war against the Sri Lankan government, the Liberation Tigers of Tamil Eelam (LTTE) established foreign offices, dispatched political counselors, and engaged in active lobbying in a number of states.3 The rebel group further operated a foreign headquarters, called Eelam House, located at 202 Long Lane in London, to serve as a clearinghouse for all LTTE overseas political activity.4 In El Salvador, the Farabundo Martí National Liberation Front (FMLN) created a Political Diplomatic Commission—its own foreign ministry—in 1981 in the midst of its civil war. The Commission posted scores of rebel “ambassadors” around the world to shore up support from governments, political parties, and international organizations.5 In the early 1990s, the Kurdistan Front office in London “played the media buzz for all it was worth” while its representatives made appearances at policy think tanks in Washington and New York.6 In the Congolese civil war of the late 1990s, the Congolese Rally for Democracy (RCD) hired lobbying firms to make its political appeals in Washington.7 Even in a “forgotten struggle” that was the Free Papua Movement (OPM), fighting for secession from Indonesia in the 1960s, the rebels opened offices in Dakar and The Hague in an effort to gain international recognition.8 And lest these examples give the impression that rebel diplomacy is a contemporary phenomenon, the Continental Congress dispatched representatives such as Benjamin Franklin, John Adams, and Thomas Jefferson across Europe to seek aid and recognition during critical phases in the American Revolution.9 Likewise, during the American Civil War the Confederacy dispatched “commissioners” to Paris and London to shore up external support.10

On the face of it, this flurry of rebel diplomatic activity across time and conflicts may be unsurprising. Many rebel groups seek and receive foreign support during civil war; dispatching representatives, opening offices abroad, lobbying in foreign capitals, and creating foreign affairs departments would seem logical steps toward gaining international sympathy. At the same time, the great lengths to which many rebel groups go to present a political face to the world in the midst of a bloody conflict begs an explanation. The puzzle here is twofold. First, studies of external intervention in civil wars show that foreign governments provide support based on their own strategic calculations, little of which have to do with how diplomatically active a rebel group has been. Why, then, would rebels shuttle around the world, when they could be focusing on making military and political gains right at the front lines?11 And second, why do some rebel groups conduct active diplomacy in wartime whereas others do not?

I define “rebel diplomacy” as a rebel group's conduct of foreign affairs during civil war for the purpose of advancing its military and political objectives. It should be understood as a rebel group's wartime political tact—as “rebelcraft”—in the same way that diplomacy is a form of statecraft in international relations. I limit the focus of this study to rebel diplomacy aimed at securing political or material advantages for the rebels' armed struggle, leaving aside diplomacy aimed at negotiating an end to the conflict. In focusing on rebels' overseas diplomatic engagements, this article offers an analysis of one prominent way in which violent nonstate actors seek their standing in international politics. I show that rebel diplomacy is a common feature of contemporary civil wars. Because rebel organizations seek formal state power and because states operate in an international system, the former are concerned not only with local battlefield outcomes but also with the international politics of a civil war. Diplomacy represents one dimension of a broader effort by rebel groups to attain visibility, credibility, and acceptance on the world stage. Through it, rebel groups aim to signal to international audiences that they are serious political contenders for state power, can adopt state-like behavior, are amenable to peaceful talks, and champion causes that may have wider international appeal.

Yet, not all rebel groups place equal value on diplomatic engagement. In what is to my knowledge the first systematic cross-national analysis of rebel diplomacy, I show that rebels for whom domestic and international political support is particularly crucial for the attainment of their goals—secessionist groups and groups concerned with domestic political organization—are more likely to become wartime diplomats. Although one might expect militarily weak or resource-poor rebels to resort to diplomacy, I find no evidence to support this notion. This finding suggests that rebel diplomacy is not a weapon of the weak, but a tactical choice for rebel groups seeking political capital within an international system that places formidable barriers to entry for nonstate entities.

The study of rebel diplomacy holds important implications for both scholarship and policy. First, it provides insights into the broader interplay between violent rebellion and international politics. While existing studies show that internal conflicts often have significant international dimensions—as put into stark relief by the ongoing conflicts in Syria, Iraq, Afghanistan, and Libya—most focus on the decision calculus of external states, especially when they choose to intervene in civil wars and through what means.12 But states alone do not shape internal conflicts, and rebel groups are not mere passive recipients of assistance. In examining rebel diplomacy, this article takes squarely into account the agency of rebel groups and their strategic calculations about whether and how to engage with external politics. More broadly, the study aims to establish the international relations of rebel groups as an important research agenda. Rather than the outside looking in, the analysis here has internal nonstate actors looking out at a state-centric world, trying to gauge how to act.13

Second, the study of rebel diplomacy adds an international dimension to the growing literature on rebel organization and governance, which shows that rebel groups engage in a gamut of activities intended to cast themselves as political organizations that are capable of governing territories and populations, creating order, and administering laws in the midst of armed conflict.14 This literature has so far paid scant attention to rebels' international activism, this aspect of warfare having been overshadowed by a dominant scholarly focus on rebels' domestic organization, local governance, and violent strategies. This article shows that many rebel groups demonstrate as much concern for overseas diplomatic campaigns as they do for domestic and local political organization. Just as through local governance rebels seek to secure popular support, so through diplomacy rebels seek external support that would help them advance their cause on the global stage. Civil wars may be fought internally, but any group fighting for state power must also navigate the international political system within which states operate.

Third, the study of rebel diplomacy provides a more complete profile of violent and nonviolent instruments available to rebel groups. Recent studies have examined a range of political tactics adopted by rebel groups as part of their armed campaigns, including civilian governance, propaganda, alliance formation, and compliance with international law.15 Further, studies show that many of these tactics are directly linked to their violent strategies. Likewise, the study of rebel diplomacy is important on its own terms as a form of rebel engagement with external actors, but it can also be extended to new analyses of the ways in which rebels' international political interests might affect locallevel patterns of violence.

In terms of policy, if diplomacy is a tool used by rebel groups seeking domestic and international support, then external states may be able to leverage the offer of diplomatic engagement to influence the behavior of such groups. Diplomacy is a form of communication, but states and international organizations can use the promise of diplomacy itself to shape rebel behavior, quite apart from the content of that communication. Although diplomatic interaction with rebel groups can be a politically sensitive issue for third-party states, it is less politically costly than the offer of overt military or financial support. It can be used to pressure rebel groups toward the lawful treatment of civilians or to keep their behavior otherwise aligned with the strategic interest of the external patron. Furthermore, this study can shed light on which types of rebel groups are likely to be more responsive to diplomatic overtures. Groups that demonstrate concern for domestic popular mobilization and those that deem international support indispensable to success are more likely to respond positively to offers of diplomatic engagement. In contrast, those that show few such concerns may require external states to employ other instruments of statecraft, such as sanctions or military support for the government fighting against the group.

The next section uses a new dataset of rebel diplomacy and its qualitative supplement to provide a descriptive account of the phenomenon in the 127 major civil wars that ended between 1950 and 2006. I then lay out a theory of rebel diplomacy and identify testable hypotheses on why some rebel groups conduct diplomacy and others do not. Using the dataset, I conduct statistical tests of the hypotheses while presenting accompanying examples from a range of cases. For further descriptive and causal analysis, the ensuing section examines rebel diplomacy of the National Union for the Total Independence of Angola (UNITA) in the 1975–2002 Angolan civil war. Within this case, I leverage variation in UNITA's diplomacy over time, using both archival and secondary sources. I conclude with a discussion of the theoretical and practical implications of this work for understanding rebel behavior and conflict dynamics.

## Rebel Diplomacy since 1950

How common is rebel diplomacy in contemporary civil wars? To answer this question and to enable both descriptive and causal analysis, I systematically collected data on rebel diplomatic activities for the main rebel group in each of the 127 major civil wars ending between 1950 and 2006.16 Actors engage in diplomacy in a variety of ways, as discussed below. The operationalization of rebel diplomacy in this study parallels states' practice of traditional (as opposed to public or cultural) diplomacy as commonly understood. Specifically, for the purpose of this study a rebel group conducts diplomacy when it engages in any of the following acts during an armed conflict against the state: (1) opens a political office abroad; (2) sends representatives abroad on political missions; or (3) creates a political body devoted to the conduct of foreign affairs. This operationalization identifies rebel groups that demonstrate their commitment to, and investment in, managing foreign affairs; it helps to distinguish them from groups that may engage in propaganda or strategic talk but which fall short of these clear indications of intentional diplomatic engagement. To collect the data, I conducted an extensive search of open sources for each civil war, in addition to consulting with area experts for certain cases. To maximize transparency and facilitate future research, a qualitative supplement to the data documents all sources used to arrive at each coding decision, as well as coding notes and direct excerpts of relevant texts.

There are limitations to the data. As is clear from other existing civil war datasets, many conflicts have involved multiple rebel groups fighting against the state.17 The dataset used here includes information only on the main rebel group in each conflict.18 Furthermore, the data impart no information about variation in diplomacy over time within each rebel group. These limitations stem from the difficulty of collecting data on rebel diplomacy, which necessarily involves consulting multiple sources for indications of the relevant activity. Even where basic facts are available, further details—such as the precise mission of a foreign office, its dates of operation, the nature of its staffing or funding, its mode and frequency of communication with the rebels' core leadership, and whether the office had been extended a formal invitation by the host state or instead operated in the shadows—tend to be paltry.

Several consequences follow from these limitations. First, the data may be selecting on rebel group strength: it is those groups that rose to relative prominence and managed to inflict greater casualties that are included in the dataset, whereas many smaller rebel groups are omitted. This can bias analysis if group strength is correlated with an independent variable, such as rebel groups with a legal political body. Second, even for those rebel groups that are included in the dataset, the scarcity of documentation on rebel diplomacy may have given rise to false negatives.19 Again, bias can result if the false negatives systematically correlate with some feature of the conflict, such as its scale or amount of media reporting. Third, because the data do not code variation in diplomacy within conflicts over time, they are unable to assess whether certain wartime covariates, such as rebels' battlefield losses, affect their diplomatic behavior. Finally, the focus on the main rebel group in each war precludes an analysis of whether competition among multiple rebel groups might affect diplomacy.

Despite these challenges, there is both sufficient documentation and sufficient variation on rebel diplomacy among groups to lend confidence that the variation is real, and not an artifact of data availability or coding biases. The data limitations necessarily make this a study of diplomacy among major rebel organizations. But even among this set, the discussion below shows that some rebel groups engaged in diplomacy and others did not, and that among those that did, rebels used diplomacy in varied and often inventive ways. The puzzle motivating the study therefore stands, and the empirics offered here can begin to provide answers. Where feasible I report ways some of the biases noted above might have affected findings.

Turning to the data itself, I find that diplomacy has been a common form of rebelcraft in the major civil wars ending between 1950 and 2006: 39 percent of rebel groups engaged in diplomacy, as operationalized. The range of rebel groups that used diplomacy is notably diverse. The Mozambican National Resistance (RENAMO) reportedly maintained offices in Heidelberg, Lisbon, Nairobi, and Washington.20 It further boasted an external relations department and had a secretary for foreign affairs who resided in West Germany.21 The Front for the Restoration of Unity and Democracy (FRUD), active in Djibouti in the early 1990s, sent its leaders abroad to meet with various heads of state, in addition to operating an office in the Yemeni capital, Sanaa.22 The Zimbabwe African National Union (ZANU) had Herbert Chitepo, the chairman of its high command, take a series of trips across the socialist world, including Bulgaria, Romania, and Yugoslavia, as well as to Australia, India, the Netherlands, New Zealand, and Sweden.23 ZANU offices in London, New York, and Stockholm were actively engaged in garnering supporters.24 The Maoists of Nepal, the Workers' Revolutionary Party of Argentina during the “Dirty War,” the Revolutionary Front for an Independent East Timor (FRETILIN), the Fedeyeen of Jordan, and the Dhofari rebels of Oman all had foreign affairs departments operating under various names.25 The purpose of the Nepalese Maoists' “International Department,” for instance, was to expand party organization by recruiting expatriates and establishing international contacts, as well as fundraise, purchase weapons, and arrange training.26

Qualitative evidence shows that rebel groups use a number of channels to facilitate diplomatic engagement. One common tactic is to hire a public relations (PR) firm to stage-manage the group's appearances abroad. The Biafran secessionist rebels in Nigeria had a Geneva-based PR firm called Mark Press, and the Sudan People's Liberation Army (SPLA) a Washington-based firm, each working to further the respective rebel group's cause abroad.27 Another common diplomatic tactic is to make use of a rebel group's diaspora population to run foreign offices and lobby foreign legislatures, as did the Free Aceh Movement (GAM) of Indonesia and the LTTE.28 A less common form of diplomacy is for a group to create its own international organization in an effort to gain prominence and strengthen solidarity. The Sikhs of India did this in 1987 in forming the Council of Khalistan, which was to serve as the Sikh rebels' foreign policy organ.29

## Explaining Rebel Diplomacy

Why do rebel groups conduct diplomacy? And why do some groups do so whereas others do not? Given that rebel groups are usually at a resource disadvantage relative to states,30 it is possible that groups that are particularly resource-poor or militarily weak relative to the state are more likely to conduct diplomacy. I visit this hypothesis based on rebels' material needs below. But given the particularities of the international system, I propose that there is an important political logic to rebel diplomacy.

### THE INTERNATIONAL POLITICS OF REBELLION

Key to understanding rebel behavior beyond the battlefield is the fundamental notion that the military contest is only one dimension of a civil war. Armed confrontations do not take place in a vacuum, but rather in specific domestic and international political contexts, and these contexts have significant bearing on war dynamics and outcomes. Rebel groups are political actors who seek the status and power of states, whether by taking over the capital or founding a new state of their own. As such, they wish to be taken seriously as political entities that are worthy of formal status in international politics—they seek de jure, not just de facto, state authority.31 Thus, as they fight militarily, rebel groups face strong pressures to simultaneously fight a political contest over allies, endorsement, and legitimacy both at home and abroad. The contest, furthermore, is often a zero-sum game: an external state's offer of material aid to the rebels is typically an act of open hostility against the incumbent regime; a state's decision to confer formal recognition on the rebels is a decision to delegitimize the incumbent government entirely.

The stakes of this contest are high, as the attainment of international support can confer on the rebel group a range of benefits. The most significant from an international political and legal standpoint is the granting of formal recognition. Should a sufficient number of states, including, crucially, the great powers, grant a rebel group recognition, the group can obtain legal state status in international politics and all its attendant benefits, such as representation in world bodies and capitals, access to global markets, foreign aid, and official bilateral relations.32 Although shy of achieving formal state status, Polisario, a group fighting for the independence of the Sahrawi Arab Democratic Republic from Morocco, has, through both violent and nonviolent means, secured diplomatic recognition from scores of states in Africa and beyond.33 Political support may also come in the form of the granting of membership in an international body. The Organization of the Islamic Conference recognized the Moro National Liberation Front of the Philippines as the “sole legitimate representative of the Bangsamoro people” in the late 1970s and granted it observer status in the body.34 Although the practical benefits of such offers may be small, these acts of political recognition by international actors can have major symbolic significance for nonstate entities—they indicate that the group is one step closer to formal acceptance into an international system that heavily guards against entry by nonmembers. Political support may certainly lead to the offer of material aid for the ongoing con°ict, such as funding, weapons, logistical assistance, training, intelligence, sanctuary, and troops. It may also come in the form of expressions of sympathy without overt promises of assistance. In international politics, the perception of political legitimacy—the idea that a group is a viable political entity—can be consequential for rebel groups' survival and success.35

All told, should a rebel group obtain preponderant international support for its cause, it can overcome even severe limitations in its military capability vis-à-vis the state and achieve its ultimate political objective. As it fought a decades-long guerrilla war against the formidable Indonesian military, for instance, the Timorese resistance successfully marshaled overwhelming international support for secession and secured East Timor's independence in 2002—an unthinkable turn of events if civil war outcomes were determined through military contests alone. Such a configuration of the international system helps explain why the U.S. State Department's conferral of foreign mission status to the Syrian National Coalition in 2014 was politically significant for the latter, even if its practical benefits were rather slim: according to a Coalition representative, it was a “clear sign” of legitimacy from the United States.36 Appealing to the political interest of potential external backers is therefore a first-order concern for rebel groups seeking any combination of the bundle of goods—formal recognition, money, weapons, international organization membership, political sympathy, and so on—that can come with international support.

For rebel groups, furthermore, the need to campaign for international support can often surpass the need to attract domestic support. Domestically, much of what a rebel group might seek in wartime, such as war taxes, territory, intelligence, or logistical aid from local populations, could be attained as much through coercion as through persuasion—in fact, rebel groups usually use a combination of the two.37 Popular legitimacy could facilitate the attainment of these goods, but it is not a prerequisite because rebels could resort to coercion and extortion.38 In contrast, in the state-centric international system rebel groups lack the power to forcibly extract support from external states.39 Instead, they need to convince international players to come to their aid through both words and actions. More than in domestic politics, then, in international politics rebel groups have incentives to watch what they do and say.

### THE FUNCTIONS OF REBEL DIPLOMACY

Given the importance of seeking international support, wartime diplomacy helps rebel groups enhance their political appeal in several ways. First, diplomacy allows rebel groups to present a political face to the international community. Through diplomacy, rebels can assure states that they are more than a violent military organization, are politically organized and willing to cooperate, can mingle with policymakers and diplomats, and are capable of holding peaceful talks. Diplomacy, in this sense, is consistent with other forms of rebelcraft, such as rebels operating their own radio stations, creating their own flags and insignia, or, more recently, establishing an online presence via group websites and social media such as Twitter.40

Diplomacy, however, is not mere show; it is also about substance. Rebel groups can use diplomacy—and the crafted rhetoric that buttresses diplomatic engagements—to persuade actual or potential foreign sponsors that the group espouses goals and interests that align with theirs and that supporting the group is in their best strategic interest. Diplomacy is a way for rebels to communicate, sincerely or otherwise, their intentions and preferences to a target audience and thereby win its support. Thus, East Timor's external delegation at the United Nations made concerted efforts to persuade sympathetic states to support first formal discussions of, then a move toward, the independence of East Timor from Indonesia.41 Once political support has been obtained, rebel groups that seek material aid can use diplomatic platforms to make direct appeals for specific wanted items, as the Syrian Coalition leaders did in requesting antiaircraft weapons as they made their rounds in Washington.42 Rebel groups are self-consciously strategic in how they “market” themselves to potential supporters.43 Diplomacy is one way through which rebel groups attempt to disseminate tailored messages, shape their international image, and ultimately win credibility and support on the world stage.

Second, diplomacy is what states “do.” As actors vying to control a state, rebel groups wish to show that they, too, are capable of conducting an important act—foreign affairs—that is typically understood to be the preserve of recognized states. Indeed, beyond diplomacy, a broader argument can be made that support-seeking rebel groups will emulate all sorts of behavior associated with “good citizenship” in the international system. For instance, in the realm of international law rebel groups sometimes go out of their way to showcase their compliance with international humanitarian law.44 In the realm of rhetoric, rebels often readily, and strategically, adopt the international jargon of the day. The Dhofari rebels of Oman changed their group name from the Dhofar Liberation Front to the Popular Front for the Liberation of the Occupied Arabian Gulf in 1968 to signal a closer alignment with the communist bloc;45 Acehnese rebel leaders, upon entering the peace process, “borrowed wholesale” the jargon of international development agencies and made commitments to a cleaner government and transparency.46 In the realm of institutions, rebels' wartime governance of civilians—reflected in the establishment of rebel schools, hospitals, laws, and courts—can be seen both as an attempt to control a population and attract its support and as an instance of deliberate mimicry of the basic institutions associated with statehood. Likewise, externally, rebels conduct diplomacy out of a calculated belief that aspiring states can reap benefits by adopting state-like behavior.

Governments fighting against violent oppositions are often themselves conducting extensive diplomacy in efforts to secure foreign allies. Rebel diplomacy is therefore often counter-diplomacy aimed at discrediting the state against which the group is fighting, a political battle fought in the international arena. In the Democratic Republic of Congo, the government of Joseph Kabila as well as the Rally for Democracy rebels both “embarked on a race for international support that clearly demonstrated the superseding of domestic agendas by strategies of extraversion.”47 When civil war broke out in Libya following the Arab uprisings of early 2011, representatives of both Muammar al-Qaddafi's government and the opposition Transitional National Council shuttled around Europe with the aim of securing or keeping external support. The intensity and significance of this political contest magnified once France, and then a host of other states, began to recognize the Council as Libya's legitimate government.48

Third, diplomacy helps support-seeking rebels by enabling them to boost their image domestically. Even if the primary targets of diplomacy are the rebels' actual or potential foreign patrons, rebels' international diplomatic activism can have feedback effects domestically. Rebel diplomacy can attract headlines. It imparts the message that the group is politically active not only at home but also internationally, that it can secure tête-à-têtes and photo-ops with foreign heads of state and maintain overseas branch offices, and, crucially, that the group is not fighting alone but has the backing of foreign allies. Because rebel groups must fight against the state for popular loyalty and support and prevent defections, these overseas feats should serve them domestically as well. Dennis Tull goes as far as to posit, in the Congolese setting, that outreach to external states is “nowadays a distinct precondition to acquire and maintain domestic constituencies.”49 Domestic legitimacy, in turn, can have its own feedback effect on rebels' international politics, enhancing their image abroad as a group that enjoys local support.

### EXPLAINING VARIATION IN REBELS’ DIPLOMATIC ACTIVISM

Given that international support can confer a range of political and material benefits on rebel groups and that diplomacy is one means of seeking it, what explains why some rebel groups become active diplomats in wartime and others do not? There is significant evidence to suggest that rebel groups place different value on the pursuit of international support.50 Following the political logic of diplomacy, one would thus expect those groups with greater concern for winning the political contest for international support and its attendant benefits to be more likely to conduct diplomacy. This expectation leads to four testable hypotheses.

First, a number of recent works show that a particular type of rebels—secessionists—have the greatest incentive to pursue the most extensive and politically significant type of status to which a nonstate actor could aspire: formal recognition.51 For secessionist groups, international legitimacy, granted through de jure recognition of the rebels by the great powers, is a sine qua non for success; without formal recognition, these rebels will be unable to achieve the ultimate goal of statehood.52 Studies show that these groups attempt to boost the perception of legitimacy among powerful states through strategic battlefield choices: secessionists are less likely to use violence against civilians, less likely to use child soldiers, and more likely to allow the International Committee of the Red Cross access to detainees.53 Beyond their military tactics, I hypothesize that secessionists are also more likely to engage in overseas activism in the form of diplomacy. Nonsecessionists may certainly seek international recognition when recognition itself is contested, as illustrated by the Syrian and Libyan examples above and the case study of Angola's UNITA, below. But across the universe of cases, one should expect secessionists to have stronger incentives to bring their cause to the international stage, for the simple reason that independent statehood is unobtainable without explicit endorsement by the great powers and a great number of other states. By allowing rebel groups to exhibit state-like behavior, present a political face, network with powerful leaders, and make the case for their cause, diplomacy can serve as a critical tool in the politics of secessionism. Thus in Sri Lanka, the LTTE, fighting for an independent Tamil nation, engaged extensively in wartime diplomacy to seek overseas support while the Janathā Vimukthi Peramuṇa (JPV), violently fighting for formal participation in national politics during much of the same period, did not, choosing instead to focus on urban insurrections supported by unemployed youth.54 If the international system is a social place, with explicit and implicit codes of “acceptable” behavior,55 secessionist rebels should be particularly keen to use diplomacy to publicize internationally that they have the competence to assume statehood and are hence worthy of membership in the exclusive club of recognized states.

H1: As groups seeking formal international recognition, secessionist rebel groups are more likely to engage in diplomacy than are nonsecessionist groups.

Second, rebel groups that demonstrate concern for domestic political organization may be more likely to engage in diplomacy. This hypothesis rests on a two-sided logic. On the one hand, seeking external support often requires rebel groups to demonstrate their domestic political competence. Studies show that rebels often undertake measures, such as compliance with international humanitarian law, not as part of a coherent military strategy but to please external audiences.56 One the other hand, rebel groups seeking domestic support often do so not only by demonstrating that they can bring political order and govern populations locally, but also by taking their political missions abroad in efforts to signal to local constituents that the group has acquired external states' support. This appears to have been the strategy of ZANU. The group built up a sophisticated political organization and governance structure locally while conducting extensive overseas diplomacy, realizing that it needed to secure both empirical sovereignty at home and juridical sovereignty abroad and that the two fronts worked in tandem.57 When the humanitarian wing of the Eritrean People's Liberation Front distributed relief supplies to people within their territories, the rebel group made sure to broadcast to beneficiaries that these supplies came from Western donors: doing so could provide the people a “psychological boost” from knowing that their fight for Eritrean independence had international attention while also giving the rebel group itself a boost of political legitimacy.58 Domestic political organization and diplomatic activism can thus comprise a two-pronged strategy of violent rebellion, with mutually reinforcing feedback effects between the two realms. In contrast, the Revolutionary United Front of Sierra Leone, a group characterized by the lack of concrete programs of social change and efforts to win popular support, demonstrated no interest in wartime diplomacy. Its predominant tactic, instead, was terror campaigns against local civilians.59

H2: Rebel groups that engage in domestic political organization are more likely to engage in diplomacy.

Third, rebel groups fighting against more repressive governments may have greater incentive to use diplomacy to gain international support, for at least two reasons. First, they may turn to diplomacy because successful domestic political mobilization is difficult, given multiple constraints in highly autocratic settings. In the conflict in the South Moluccas in Indonesia in the 1950s, rebel leaders believed that international diplomacy would be more effective than negotiations with the central government.60 Likewise, early in the Acehnese conflict, leaders of the Free Aceh Movement saw “internationalization … as the only way to level the playing field with Indonesia.”61 Where domestic obstacles to successful rebellion are paramount, actors may turn overseas. Second, rebels in these states may be more inclined to use diplomacy to delegitimize the incumbent regime. Genuine or not, the rhetoric of violent rebellion has been replete with references to peace, liberation, democracy, freedom, and elections; as mentioned, rebel groups' diplomatic messages often directly serve as rhetorical weapons against the incumbent regime. Diplomacy can thus become a tool in an international delegitimization campaign against the home government, and the impetus for such campaigns is likely to be greater in authoritarian contexts.

H3: Rebel groups fighting against more authoritarian regimes are more likely to engage in diplomacy.

Finally, apart from the political logic laid out above, there may be a straightforward material logic to rebel diplomacy. If one motivation for rebels to prove their worth internationally is to secure military or financial support from patrons, then rebel groups with greater need for material aid should be more likely to engage in diplomacy. It is possible that for such rebels, diplomacy is mere “cheap talk”—that is, a way to simply achieve the ultimate goal of material support. The need for external resources may be most acute for those groups whose military capability vis-à-vis the state is particularly weak. If prospects on the battlefield are bleak, rebel groups should be especially keen to turn to external patrons for warfighting resources. Diplomacy, in this sense, may be a favored weapon of the weak. Militarily strong rebels, in contrast, should be able to garner external support without having to appeal for it, because states are more willing to sponsor such rebels to begin with.62

H4: Rebel groups that are weaker relative to the state are more likely to engage in diplomacy.

## The Role of External Supporters

The theoretical discussion above focuses almost entirely on rebel incentives. Diplomacy, however, involves interaction between at least two parties. Although rebel groups may create ministries of foreign affairs at will, opening consular offices and making high-level visits to foreign capitals would seem to require that the host states abroad first extend their hands to the rebels. Yet, there are empirical and theoretical reasons why an explanation for rebel diplomacy can stand without explicitly incorporating the incentives of foreign host states. First, empirically, rebel diplomacy can and does occur without formal invitations from abroad. When rebel leaders meet with foreign heads of state and make headlines, what is being observed is typically the culmination of years of diplomatic groundwork in which rebel leaders tour abroad to expand their often very personal network of sympathizers. In the Acehnese conflict, for example, GAM rebel leader Hasan di Tiro sought over several decades to foster international support for Acehnese independence, starting with his American contacts from his graduate school days.63 After years of feeble support, a breakthrough came around 1985 when di Tiro met the Libyan ambassador to Sweden, an old acquaintance from his time in business. The ambassador quickly arranged for di Tiro to visit Libya, where he secured Muammar al-Qaddafi's agreement to train the Acehnese fighters.64 In other words, diplomacy had been from the outset a core part of GAM's strategy of rebellion, long before heads of state, embassies, the United Nations, and other international actors began to make “official” contacts with the group. Likewise, in the Namibian conflict members of the South West Africa People's Organization (SWAPO) were positioned in the United States, various parts of Europe and Africa, and the United Nations, and “were able to establish, most often from scratch, effective networks of contacts and support in the countries in which they worked.”65 Rebel representatives were thus conducting diplomacy long before external states reciprocally showed interest in formally supporting the group.

The overseas posting of rebel representatives aside, what of the opening of rebel offices abroad? Surprisingly, this, too, often occurs through “unofficial” channels—that is, without formal invitations—as the Angolan case study below discusses in depth. Certainly, rebel “consular” offices may at times enjoy official status as such in friendly states, as with the Syrian National Coalition offices not only in Washington and New York but also in Ankara, Berlin, London, Paris, and elsewhere.66 More often than not, however, and especially in Western democracies, rebel offices easily register as quasi-think tanks, lobbying groups, youth organizations, and cultural centers, thus bypassing any need for formal invitations from host states.

Theoretically, thinking about rebel diplomacy as an interplay between rebels' and host states' political incentives in fact confounds the research question, turning it into one about effectiveness. When rebel diplomacy is successful, external states respond by extending the rebel group invitations to open embassies and meet with heads of state, granting recognition to the group, and so on. This study, however, is about why rebels conduct diplomacy, not about the effectiveness of their efforts. Just as diplomatic success can motivate further diplomacy, so might diplomatic failures. Examining the behavior of external states vis-à-vis rebel diplomats is therefore interesting, but neither reveals the full extent of rebel groups' diplomatic activism nor is theoretically necessary for explaining why some groups conduct diplomacy and others do not. When strategically meaningful, rebel groups will attempt diplomacy regardless of whether a formal invitation has been extended by external states.

## Data and Research Design

Having provided a range of examples and some initial evidence of the causal arguments, in this section I use statistical analysis based on cross-national data to conduct a systematic test of the hypotheses. Above, I described the dependent variable, rebel diplomacy, a dichotomous variable that has been coded for each of the 127 major civil wars ending between 1950 and 2006. The independent and control variables are summarized in table 1 and described below.

Table 1.

Summary Statistics of Independent and Control Variables

VariableMeanStandard DeviationMinimumMaximumSources
Secessionist rebels 0.35 0.48 Fearon & Laitin 2003, extended by Fortna & Huang 2012
Schools 0.30 0.46 Huang, forthcoming
Legal political wing 0.20 0.40 NSA dataset (Cunningham, Gleditsch, & Salehyan 2013)
Prewar Polity score –3.18 5.29 –10 10 Polity IV (Marshall & Jaggers 2005)
Mobilization capacity 0.65 0.48 NSA dataset
Contraband 0.15 0.36 Fortna 2008
External support 0.61 0.49 NSA Dataset
Cold War conflict 0.68 0.47 Doyle & Sambanis 2006, UCDP/PRIO Armed Conflict Dataset v. 4 (Gleditsch et al. 2002)
War duration (In) 3.69 1.45 6.17 Doyle & Sambanis 2006, UCDP/PRIO Armed Conflict Dataset v. 4
Marxist rebels 0.24 0.42 Kalyvas & Balcells 2011
VariableMeanStandard DeviationMinimumMaximumSources
Secessionist rebels 0.35 0.48 Fearon & Laitin 2003, extended by Fortna & Huang 2012
Schools 0.30 0.46 Huang, forthcoming
Legal political wing 0.20 0.40 NSA dataset (Cunningham, Gleditsch, & Salehyan 2013)
Prewar Polity score –3.18 5.29 –10 10 Polity IV (Marshall & Jaggers 2005)
Mobilization capacity 0.65 0.48 NSA dataset
Contraband 0.15 0.36 Fortna 2008
External support 0.61 0.49 NSA Dataset
Cold War conflict 0.68 0.47 Doyle & Sambanis 2006, UCDP/PRIO Armed Conflict Dataset v. 4 (Gleditsch et al. 2002)
War duration (In) 3.69 1.45 6.17 Doyle & Sambanis 2006, UCDP/PRIO Armed Conflict Dataset v. 4
Marxist rebels 0.24 0.42 Kalyvas & Balcells 2011

SOURCES: James D. Fearon and David D. Laitin, “Ethnicity, Insurgency, and Civil War,” American Political Science Review, Vol. 97, No. 1 (February 2003), pp. 75–90; Virginia Page Fortna and Reyko Huang, “Democratization after Civil War: A Brush-Clearing Exercise,” International Studies Quarterly, Vol. 56, No. 4 (December 2012), pp. 801–808; Reyko Huang, The Wartime Origins of Postwar Democratization: Civil War, Rebel Governance, and Political Regimes (Cambridge, : Cambridge University Press, forthcoming); David E. Cunningham, Kristian Skrede Gleditsch, and Idean Salehyan, “Non-State Actors in Civil Wars: A New Dataset,” Conflict Management and Peace Science, Vol. 30, No. 5 (November 2013), pp. 516–531; Monty G. Marshall and Keith Jaggers, Polity IV Project: Political Regime Characteristics and Transitions, 1800–2004 (Arlington, Va.: Center for Global Policy, George Mason University, 2005); Virginia Page Fortna, Does Peacekeeping Work? Shaping Belligerents' Choices after Civil War (Princeton, N.J.: Princeton University Press, 2008); Michael W. Doyle and Nicholas Sambanis, Making War and Building Peace: United Nations Peace Operations (Princeton, N.J.: Princeton University Press, 2006); Nils Petter Gleditsch et al., “Armed Conflict 1946–2001: A New Dataset,” Journal of Peace Research, Vol. 39, No. 5 (September 2002), pp. 615–637; and Stathis N. Kalyvas and Laia Balcells, “International System and Technologies of Rebellion: How the End of the Cold War Shaped Internal Conflict,” American Political Science Review, Vol. 104, No. 3 (August 2010), pp. 415–429.

NOTE: NSA stands for the Non-State Actors in Armed Conflict Dataset, and UCDP/PRIO stands for Uppsala Conflict Data Program/Peace Research Institute Oslo.

To test H1, I use a dichotomous variable for “secessionist rebels,” which distinguishes rebel groups that sought territorial autonomy or independence from those that aimed to take over the central government. Thirty-five percent of rebel groups in the dataset had separatist aims, whereas 65 percent aimed for regime overthrow. I use two alternative measures of rebel groups concerned with domestic political organization and support (H2) to ensure results are not variable-specific. The first is a dichotomous variable for whether or not a rebel group established its own schools in the midst of conflict. The variable is taken from the Rebel Governance Dataset, which contains data on rebel groups' wartime governance of civilians and funding sources for the major post-1950 civil wars.67 Studies show that rebel groups provide social services such as education to demonstrate their commitment to civilians, secure their consent, and achieve a degree of popular legitimacy.68 According to the data, 30 percent of rebel groups—including the Tigrayan People's Liberation Front of Ethiopia, the Pathet Lao of Laos, and Polisario of Western Sahara—opened and operated their own schools during conflict. The second measure for domestic political organization is a dichotomous variable for rebel groups with a legal political wing (20 percent of the cases), taken from the Non-State Actors in Armed Conflict Dataset (NSA).69 Having a political body that has been granted legal status in the state should be a strong indicator of rebel groups seeking domestic organization and even participation in formal politics.

To measure the degree of autocracy of the regime in power in the conflict state (H3), I take the average of the state's Polity IV scores from the three years prior to the start of the war. For rebels' relative capability vis-à-vis the government (H4), I use a measure for the rebel group's ability to mobilize popular support relative to the government. The variable is coded 1 if mobilization capacity is coded “moderate” or “high” in the NSA, and 0 if coded “low.”70 As the authors of the NSA data point out, rebel groups that can mobilize greater numbers of personnel should pose greater military threats to governments.71 To ensure results are not dependent on this particular measure, in alternative models I also include two variables that jointly measure rebels' financing capability, which is arguably a significant dimension of rebels' overall capacity to fight. These are rebels' dependence on contraband and rebels' receipt of foreign material support.72 Rebel groups with revenue from the illicit extraction or sale of drugs, minerals, and other resources or from external sponsors should have greater capacity to fight compared to those without these sources of war-financing.73 Inclusion of the foreign support variable also addresses the concern that diplomacy may be contingent on the rebels already having received some external support, as well as the possibility that some of the other independent variables, such as rebel schools, may be endogenous to rebel groups' overall capability.

The regression models include several control variables. A variable for “Cold War conflict” marks wars that began before 1989. The two superpowers' engagement in proxy wars during the Cold War decades meant that extensive material aid flowed to various states and rebel groups in accordance with their ideological leanings. Wartime statecraft and rebelcraft may have been less consequential at this time, given that the superpowers proved willing to dole out aid so long as their recipients were ideological allies. After the Cold War ended, the granting of both resources and legitimacy by external states may have become much more conditional on rebels (and states) exhibiting behavior deemed appropriate by the foreign sponsor. It is thus possible that diplomacy, as a means through which states and rebels broadcast their missions and obtain external support, has become a more critical tool in the post–Cold War period.

I also control for the “duration of the war,” counted in months and logged, as longer wars provide more time and opportunities for rebels to conduct diplomacy.74 Finally, in some of the models I control for “Marxist rebel groups” to test an alternative argument based on rebel ideology. Marxist rebel groups typically claim to be fighting a “people's war” and place importance on garnering popular support. These organizations also tend to have a clear hierarchical structure and a strong leadership. They often create “people's governments” and provide social services while emphasizing domestic and international propaganda work. Thus, it is plausible that rebels' Marxist ideology is an omitted variable driving support-seeking motives both domestically and internationally.

## Empirical Results

Table 2 presents the results of logistic regressions for the determinants of rebel diplomacy. Cases are clustered by country to account for the fact that wars in the same state may not be independent of each other. Model 1 serves as the base model; variations are introduced in models 2–5.

Table 2.

Determinants of Rebel Diplomacy, 1950–2006

(1)(2)(3)(4)
Base ModelControlling for IdeologyControlling for FinancesAlternative Measure of Domestic Legitimacy
Secessionist 0.99** 1.05** 1.05** 0.95*
(0.489) (0.507) (0.529) (0.539)
Schools 1.29*** 1.22** 1.20**
(0.479) (0.487) (0.482)
Legal political wing    1.00*
(0.542)
Prewar Polity −0.07* −0.07* −0.06 −0.05
(0.040) (0.040) (0.044) (0.043)
Mobilization capacity −0.54 −0.54 −0.50 −0.40
(0.464) (0.465) (0.477) (0.482)
Cold War −0.08 −0.16 −0.11 −0.03
(0.405) (0.401) (0.417) (0.395)
War duration (In) 0.35** 0.32** 0.28* 0.44***
(0.139) (0.160) (0.161) (0.154)
Marxist  0.35 0.39 0.82
(0.583) (0.568) (0.567)
Contraband   0.32 0.35
(0.871) (0.830)
External support   0.13 0.14
(0.516) (0.485)
Constant −2.41*** −2.31*** −2.35*** −2.91***
(0.667) (0.710) (0.728) (0.797)
Observations 126 126 126 126
(1)(2)(3)(4)
Base ModelControlling for IdeologyControlling for FinancesAlternative Measure of Domestic Legitimacy
Secessionist 0.99** 1.05** 1.05** 0.95*
(0.489) (0.507) (0.529) (0.539)
Schools 1.29*** 1.22** 1.20**
(0.479) (0.487) (0.482)
Legal political wing    1.00*
(0.542)
Prewar Polity −0.07* −0.07* −0.06 −0.05
(0.040) (0.040) (0.044) (0.043)
Mobilization capacity −0.54 −0.54 −0.50 −0.40
(0.464) (0.465) (0.477) (0.482)
Cold War −0.08 −0.16 −0.11 −0.03
(0.405) (0.401) (0.417) (0.395)
War duration (In) 0.35** 0.32** 0.28* 0.44***
(0.139) (0.160) (0.161) (0.154)
Marxist  0.35 0.39 0.82
(0.583) (0.568) (0.567)
Contraband   0.32 0.35
(0.871) (0.830)
External support   0.13 0.14
(0.516) (0.485)
Constant −2.41*** −2.31*** −2.35*** −2.91***
(0.667) (0.710) (0.728) (0.797)
Observations 126 126 126 126

NOTE: Robust standard errors appear in parentheses.

***

p < 0.01,

**

p < 0.05,

*

p < 0.1

The results provide consistent evidence that rebel groups that seek political support, whether internationally or domestically, are more likely to conduct diplomacy. Both the secessionist variable (H1) and that for rebel schools (H2) are consistently positive and statistically significant. Further analysis of substantive effects shows that a switch from a nonsecessionist to a secessionist group increases the likelihood of diplomacy by 19 percent,75 and that rebel groups that build schools are 26 percent more likely to engage in diplomacy than those that do not.76 The alternative indicator of rebels' domestic political organization, rebels with a legal political wing, is also statistically significant, as shown in model 4; such groups are 20 percent more likely to conduct diplomacy than those without.77 These findings lend support to the argument that diplomacy is a wartime tactic used by those rebel groups for whom acquiring support and endorsement, be it among local populations or among external states, is deemed critical to the attainment of their political goals.

H3 on repressive regimes finds only weak and inconsistent support in the data. The negative correlation indicates that rebel groups operating in more authoritarian states are more likely to conduct diplomacy than those operating in less authoritarian contexts, but the variable is statistically significant in only some of the models.78 This finding suggests that irrespective of the nature of grievances or level of domestic repression, rebels that seek to attract internal and external support are more likely to use diplomacy to do so.

Surprisingly, rebel groups' relative fighting capacity, measured by their ability to mobilize fighters relative to the government, has no bearing on their diplomatic engagements (H4). The coefficients are consistently negative, suggesting that weaker rebels are more likely to conduct diplomacy, as hypothesized, but the variable is not statistically significant. Models 3 and 4 include the two measures of rebels' financing capability (rebels' access to contraband and to foreign material aid for the war), which are meant to capture another dimension of rebel groups' fighting strength. Again, I find no correlation between rebels' financing capability and their engagement in diplomacy, and inclusion of the financing variables does not affect other substantive results.79 Simply put, the idea that diplomacy is a weapon of the weak finds no support in the data.80 One explanation is that there is a mutually reinforcing relationship between resource needs and diplomatic activism. Weak rebel groups with acute needs for external support may be more likely to conduct diplomacy, but once they receive aid, they may then be compelled to continue diplomatic engagements to maintain the strategic partnership. If this were the case, H4 would be supported in the early stages of a war, but in later stages resource needs would not correlate neatly with diplomacy as both the cash-strapped and the cash-endowed engage in diplomacy, though for different reasons. Longitudinal analysis would be needed to test the validity of this claim, but the case of UNITA below provides initial evidence consistent with this logic.

As for the control variables, the argument that rebels became more diplomatically active with the end of the superpower rivalry is not supported by the data. Indeed, when models 1 and 2 are run for the Cold War cases only (N=86), substantive results remain unchanged (not shown)—even during those decades rebels concerned with domestic and international political support were more likely to be diplomats. Again, the case of UNITA below, one of the most prominent proxy wars of the Cold War period, demonstrates the rebels' heavy investment in diplomacy despite the backing of a powerful ally in the form of the United States.

Models 2–5, which control for Marxist rebels, provide further evidence to support H1 and H2 while offering no support for the Marxist ideology argument. The control variable is not statistically significant, and the measures for secessionists and rebel schools remain significantly associated with diplomacy. Finally, as expected I find that rebel groups are more likely to engage in diplomacy in longer-lasting wars.81

In sum, quantitative analysis provides evidence that rebel diplomacy is driven by a political logic: groups that seek international support and its attendant benefits, such as recognition, legitimacy, material aid, and local support, are more likely to engage in diplomacy. Secessionist rebels, for whom international recognition is essential to success, as well as those that seek to organize political authority locally, are more likely to engage in diplomacy than are groups that are more narrowly focused on military gains or those with little concern for domestic organization. At the same time, diplomacy is neither a weapon of the weak nor a weapon of the heavily repressed. Regardless of their fighting capability or the nature of the state's regime, rebels are more likely to conduct wartime diplomacy if they deem political support from abroad to be especially vital to their broader campaign against the state.

## UNITA's Diplomacy in the Angolan Civil War

To complement the statistical tests of the hypotheses on the determinants of rebel diplomacy, this section engages in an in-depth study of one case, the Angolan civil war of 1975–2002. The Angolan war is well suited to both descriptive and causal analysis. First, although UNITA's diplomatic activism is unexceptional in that the rebel group is just one among many that engaged extensively in the rebelcraft, the group left a significant paper trail throughout the war. I am therefore able to use multiple sources, including scholarly studies, news articles, and archival material, to trace its diplomatic activities and conduct a descriptive analysis of an understudied phenomenon.

Second, and more important for hypothesis testing, the break in the civil war created by the peace process in 1991–92 offers two distinct cases within one long-lasting civil war. The two cases generate variation on the dependent variable: in the first phase of the war UNITA engaged in extensive diplomacy, whereas in the second phase, starting in 1992, diplomacy notably declined. I leverage this variation to conduct a causal analysis of the rebel group's use of diplomacy across time within a single conflict.82 Should my theoretical expectations hold, the positive case of major diplomatic activity should be associated with significant concerns on the part of UNITA to earn formal and informal recognition as a viable political entity both domestically and internationally; the negative case of diplomatic inactivity should correlate with an absence of such concerns and, consequently, with behavior that would suggest UNITA's disregard for external and internal support.

Note that although UNITA was not aiming for secession, its pursuit of international recognition closely resembled that of secessionist groups, for an important reason: its most powerful and valued foreign sponsor, the United States, had withheld recognition of the incumbent regime, the People's Movement for the Liberation of Angola (MPLA), as the legitimate government of Angola. As far as UNITA was concerned, formal international recognition itself could be won through victory in the civil war. This political situation allows for a test of the logic behind H1, namely, that groups seeking de jure international recognition are more likely to engage in diplomacy. The case study thus offers both an in-depth descriptive account of one group's engagement in a particular rebelcraft, as well as a causal analysis that exploits variation in the use of that rebelcraft across time.

### THE FIRST WAR: UNITA'S EXTENSIVE DIPLOMACY

The Angolan civil war began on the heels of independence from Portugal in 1975.83 It pit the MPLA government, which had seized control of Angola's capital, Luanda, in August that year, against UNITA, which sought to wrest control of the capital from it, in what became one of the most internationalized, bloodiest, and longest-running conflicts of the Cold War. The Marxist-Leninist MPLA government received significant political, military, and financial support from the Soviet Union and Cuba throughout the 1980s, while UNITA, declaring itself to be an anticommunist group, received substantial backing first from South Africa, then the United States.84

From the outset, UNITA adopted a strategy of internationalization to counter Soviet and Cuban support for the MPLA. Its diplomatic tactics were wide and varied. First, from early in the conflict the group stationed “ambassadors” in various Western capitals.85 Second, its leaders secured diplomatic passports from its friends in Africa, including Zaire, Côte d'Ivoire, Senegal, and Congo-Brazzaville.86 Third, the group opened numerous overseas offices: in addition to offices in Dakar, Geneva, Lisbon, London, Munich, Paris, and Rabat, UNITA had a full-time office in Washington.87 By the mid-1990s, UNITA maintained twelve offices abroad: two in the United States (Washington and New York), seven in Europe, and three in Africa.88

How did a violent rebel organization such as UNITA open and operate branch offices abroad, including in Western democracies? In Washington, UNITA's office was first called the Free Angola Information Service and was officially registered with the Department of Justice as a representation of UNITA.89 The office later, in the 1990s, changed its name to the Center for Democracy in Angola and, at least for a time, was headed by a U.S. national.90 In addition to representing UNITA interests in the U.S. government and among influential think tanks, the office issued regular publications relaying the rebel group's political views. The UNITA office in Portugal, going by the same name, appears to also have been run by Portuguese nationals. The German office, meanwhile, had “no official status,”91 and was entirely funded by the Hans Seidel Foundation in West Germany.92 The Belgian office operated as a cultural center; the French office, called Demain l'Angola (Tomorrow Angola), appears to have officially served as a UNITA representation in Paris; and the British office operated under the cover of an import-export company and was run by British staff.93 The formal status of UNITA offices thus varied by location, and the hiring of national leaders and staff appears to have been one strategy used to overcome the bureaucratic hurdles involved in creating these foreign branches.

Fourth, UNITA President Jonas Savimbi himself made diplomatic visits to many capitals, including six visits to Washington. On his first visit in 1979, he met with several senators and with former Secretary of State Henry Kissinger, among others. On the second, in December 1981, he met with senators, held press conferences, and attended private gatherings, along the way “charm[ing] his audiences with his speaking style and his message.”94 By the third visit, in January 1986, Savimbi's hosts were welcoming him with the kind of treatment ordinarily reserved for heads of state. He scored a meeting with President Ronald Reagan and Vice President George H.W. Bush at the White House, in addition to making appearances on prominent television programs including Nightline, 60 Minutes, and the MacNeil/Lehrer NewsHour, as well as on C-Span, a leading American public affairs television network.95 In a letter written from UNITA headquarters in Jamba, Angola, in March 1986, Savimbi thanked Vice President Bush for the “warm reception” and for “[speaking] most inspiring words about me and my people.”96 In June 1988 Savimbi toured New York, Washington, and a number of southern cities. So comfortable was Savimbi in conservative circles in Washington that by his fifth visit in October 1989, he was declaring to an audience at the Heritage Foundation that visiting the think tank was “like coming back home.”97 Savimbi's final visit to the White House took place in October 1990, when Bush expressed continued support despite mounting domestic criticism.98

A final diplomatic tactic was that, like a number of other well-financed rebel movements, UNITA relied heavily on public relations and lobbying firms to choreograph its overseas appearances. The most well-documented instance is Savimbi's 1986 visit to the United States, for which UNITA reportedly paid $600,000 to a Washington-based public relations firm called Black, Manafort, Stone, and Kelly to arrange its access to senior government officials, the public, and the media.99 Notably, in a diplomatic counterattack the MPLA government followed suit by hiring Gray & Company, another public relations firm, to orchestrate its own political maneuvers in the United States.100 Certainly, UNITA's wartime diplomacy would not have been as extensive or successful had Washington not done its own part to encourage and facilitate it. For instance, a secret 1987 White House document on its Angola policy maintained that the United States would not only ensure that its support is “responsive to UNITA's needs”; it would also “explore means of increasing UNITA's stature within Angola and internationally.”101 The “warm reception” mentioned in Savimbi's letter to Vice President Bush had been but a small part of a substantial (and controversial) U.S. backing of the rebel group through covert and overt assistance throughout the Reagan and Bush administrations. Nevertheless, UNITA consistently employed diplomacy as a part of its warfighting strategy. Although its diplomatic activism may have been in part a response to U.S. invitations and aid, UNITA had stressed the importance of diplomacy from the war's outset, even before Washington had turned into the staunch ally that it was soon to become. Why, then, did UNITA go to such lengths to conduct overseas diplomacy while fighting an intractable civil war domestically? If my theoretical argument holds, one should expect to find evidence of significant concern on the part of UNITA to gain international and domestic support and build up its political capital during this phase. Multiple sources suggest this was the case. First, the rebel group made it widely known from the war's outset that above all, it sought international legitimacy as a serious contender against the “totalitarian” MPLA regime.102 Throughout the 1970s and 1980s, UNITA consistently cast itself as a liberation movement fighting for national unity, freedom from colonialism and communism, and for “Western values of democracy, freedom of religion, freedom of speech, the right to own one's home and some land.”103 UNITA's volumes of writings, domestic and overseas speeches and statements, emphasis from its inception on political organization and development, and existing accounts of its ideology and behavior cumulatively point to the group's overriding concern to gain legitimacy and acceptance, both at home and abroad, as a political movement capable of bringing democratic governance to Angola.104 As mentioned, although UNITA was not a secessionist rebel group, the political circumstances in which it found itself arguably replicated the exigencies faced by secessionist organizations: key international players, including the United States, had withheld diplomatic recognition of the MPLA as the legitimate government of Angola, primarily because of concerns about the presence of Cuban troops in the country and their involvement in regional conflicts.105 Both the MPLA and UNITA thus saw international recognition as a prize of victory in the civil war, and this gave both parties strong incentives to appeal to external states in efforts to rally support. The Angolan case therefore lends support to the argument that it is the value rebels place on the contest for international recognition, rather than secessionism per se, that drives diplomatic activism (H1). That the MPLA government itself was conducting its own official diplomacy—touring European capitals as well as Cuba in the mid-1980s, opening its own foreign offices, and consistently engaging with Washington to gain formal recognition—added to the impetus and urgency of rebel counter-diplomacy.106 In addition to the war's military and ideological dimensions, for UNITA the diplomatic arena was another crucial battleground, one in which the belligerents wrestled for formal recognition and support from major international players. Second, UNITA had a clear, expressed desire for resources from abroad. When South African sponsorship ended in the mid-1980s, and in light of persistent Soviet and Cuban aid to the MPLA, U.S. backing became critical to UNITA's survival. Thus, as Savimbi appealed to Washington for moral and political support, he also made overt pleas for material aid, with “antiaircraft weapons, the Red-eye missile and Stinger missile, and antitank weapons” topping his list of wanted items for a time.107 Even after U.S. assistance began flowing to UNITA, the rebel leadership was aware that continued aid was far from guaranteed, given significant opposition in Congress and even within the Reagan administration.108 UNITA was not a militarily weak rebel group;109 consistent with the statistical analysis above, it was not its degree of fighting capacity, which was already fairly strong at the height of U.S. aid to the group, but rather its persistent desire for political backing and, along with it, sophisticated weapons, that motivated the rebels to use diplomacy throughout this period of the war. Consistent with the statistical results, then, the case lends no support to H4 on rebels' relative strength. UNITA's increasing military capability did not lead to any corresponding decline in its diplomatic efforts, but rather to sustained diplomacy as rebels sought to ensure continued aid. Third, UNITA used successes in international diplomacy to enhance its domestic legitimacy (H2). For all its brutality, during the height of the war UNITA demonstrated significant concern for establishing itself as a viable political organization and gaining popular support at home. A journalist visiting UNITA's “liberated” territories in 1989 wrote, “Unitaland in southern Angola is perhaps the most impressive guerrilla zone anywhere.”110 UNITA functioned as a well-organized sovereign administration in the territories it controlled, engaging with the public through the building of schools, hospitals, and roads and providing electricity, in addition to circulating its own publications, operating a radio station in several Angolan languages, and even issuing its own internationally recognized postage stamps.111 By all accounts, UNITA was a sophisticated organization with “modern bureaucratic institutions” consisting of a president; a congress; a political bureau; central, regional, and village committees; a number of ministries; and a constitution.112 According to reports, UNITA saw to it that foreign visitors recognized the full scope of these local accomplishments. Its headquarters, Jamba, “became the public relations face of UNITA shown to journalists and Western backers of all sorts flown in for a day or two and impressed by a well-supplied bush hospital, schools, a stadium, traffic lights and an airport with UNITA immigration facilities.”113 UNITA took such visitors “on tour after tour which were designed as much to show the world what they had, controlled, and could do” as they were to impress their immediate guests.114 In turn, UNITA's successive overseas achievements may have served as a significant boon to its domestic image and credibility, doing much to reassure the populace that this was not an isolated movement but one with the backing of powerful international allies. Thus, according to a 1988 journalistic account, “thousands” of prints of the photos of Savimbi with President Reagan and Secretary of State George Schultz taken during his 1986 White House visit “now cover virtually every village entrance, community center, hospital and workshop in all of what UNITA calls Liberated Angola”115—the group was clearly using its overseas accomplishments for domestic consumption. These pieces of evidence suggest that UNITA was fully capitalizing on the reinforcing feedback effects between domestic and international support. Finally, and again consistent with the quantitative analysis, the Angolan case shows that even in the context of the Cold War, diplomacy proved vitally important for the rebels' international relations. UNITA engaged in extensive diplomacy in the 1980s in part to demonstrate to Western governments that the rebels were on their side in the Cold War, and hence could be a trusted partner in fighting against the Soviet-backed MPLA. Cold War politics did not obviate the need for diplomacy; to the contrary, it all but dictated its content. ### THE SECOND WAR: DIPLOMACY SIDELINED The 1992 presidential and parliamentary elections in Angola—the first free elections in Angolan history and held in accordance with the 1991 Bicesse accords—put UNITA's declared commitment to freedom and democracy directly to the test. When the results favored the governing MPLA, UNITA, in a move signaling its abandonment of the rhetoric of democracy crafted over the preceding seventeen years, rejected the election results and quickly returned to the bush, plunging the country back into war. In response, and in the context of the end of the Cold War, a change to the Democratic administration of Bill Clinton, and UNITA's continued intransigence in peace talks, Washington quickly distanced itself from its erstwhile proxy. With the first set of UN sanctions imposed against UNITA in 1993, the shift from extolling and backing the rebel group to ensuring its political isolation seemed complete. If my theoretical logic holds, the decline in diplomacy in this phase of the conflict should correspond with evidence of diminished interest on the part of UNITA in acquiring the political benefits that could be accrued through international support. Indeed, UNITA actions at this time gave little indication that it still cared for international or domestic legitimacy or for material resources from abroad. There were overlapping reasons for the rebels' strategic reversal after 1992. First, just as the United States was withdrawing support for UNITA in the early 1990s, the rebel group turned to the mining and selling of diamonds as a substantial source of funding. By 1994 UNITA had “near total control” of the country's diamond trade,116 worth$400 million to \$600 million annually.117 Although regression analysis yielded no significant correlation between rebels' dependence on contraband and diplomacy, in the case of UNITA the newfound income from the diamond trade likely played a sizable role in decreasing the tactical utility of diplomacy. Even so, there was an even greater reason for the rebels to discount diplomacy at this time: in a major blow to UNITA, in May 1993 the United States formally recognized the MPLA as the legitimate government of Angola. The rebels had lost the battle for international recognition altogether. There are grounds to believe that this did more to convince UNITA of the futility of further diplomacy than did the new wealth from diamonds. Even as their profits from the diamond trade were rapidly increasing, on the eve of U.S. recognition of the MPLA, UNITA representatives made last-ditch attempts to dissuade the United States from taking that step, not least by writing a letter to the New York Times warning that all efforts toward peace and democracy in Angola would be frustrated should Washington use that “most potent policy tool” of recognition.118 At this eleventh hour, and despite its diamond wealth, UNITA still sought to salvage the fight for de jure legitimacy by preventing the United States from throwing its full weight behind their erstwhile shared enemy. Once the MPLA had been granted recognition, however, UNITA had few reasons to continue shuttling around the globe making entreaties for foreign aid, approval, or acceptance. Newly self-sufficient and having lost the battle for recognition, UNITA found little use for diplomacy as a warfighting tool.

Indicatively, UNITA instead turned inward at this time. It ramped up its military operations at home and, in the process, began to acquire what one scholar calls “a uniquely criminal character.”119 Whereas it had emphasized the importance of local popular support in the 1980s, once the 1992 elections results had dashed its aspirations UNITA turned against civilians and employed terror tactics such as indiscriminate killing, summary executions, and forced displacement.120 According to one observer, the situation at this time was “probably worse than at any time in the country's history.”121 Having betrayed its rhetoric of freedom and democracy and lost the battle for international legitimacy, and awash in diamond profits, UNITA no longer saw the need to invest in acquiring domestic popular support. Consistent with this logic, UNITA's civilian governance efforts also declined at this time.122 Domestic consumption, too, had ceased to be a motive for overseas diplomatic activism. Neither did UNITA make efforts to present a set of coherent political objectives in this phase, having “abandoned the strategy of creating an intimate and reciprocal relationship between the political and military aspects of the war.”123 In this phase of the war, then, UNITA had lost both its international and domestic incentives for diplomacy. Gone were the days when UNITA paid hefty sums to lobbying firms to ensure impeccable appearances on its U.S. tour. The two-pronged strategy of acquiring domestic and international political support had turned into a largely one-track military strategy of crushing government capacity and terrorizing the population.

Perhaps indicative of UNITA's crumbling control over its overseas arms, its foreign offices did not shutter entirely after 1992. In fact, while UNITA forces weakened under government attacks and its administration and leadership disintegrated, its overseas missions continued to operate relatively unscathed for a time.124 This, too, however, would change when in 1997 the UN Security Council passed a tougher set of sanctions on UNITA, one that required of all states “the immediate and complete closure of all UNITA offices in their territories.”125 Tellingly, states hosting the rebel offices faced difficulties complying with the resolution: France cited civil liberties concerns over closing down UNITA's Paris office; Germany claimed the Bonn office had no official status to begin with; Côte d'Ivoire denied having an office in Abidjan altogether; and attempts to close the offices in Washington and Lisbon were met with domestic opposition.126 Nevertheless, by 1998 most UNITA foreign offices had ceased to function.127 As for the diplomatic passports doled out by friendly African states to UNITA's senior leaders in their heyday, at least Côte d'Ivoire declared the Ivorian passports in the rebels' hands null and void in February 1999.128 UNITA diplomacy thus ended in this phase of the conflict because the rebels lost the political and financial incentives for it, and because its former patrons concomitantly took steps to sever diplomatic relations with the rebels. Diplomacy was certainly not a weapon of the weak in this case: a weakened UNITA did not increase diplomatic efforts during this phase, but instead abandoned diplomacy as a tactic.

### SUMMARY

The case of UNITA shows that when the rebel group sought international recognition, material resources, and both external and internal political support, it employed diplomacy extensively as a warfighting tactic. Diplomatic activism allowed the group to craft a tailored image for itself; spread its political messages to a wide international audience; “do” as states and statesmen do in opening foreign offices, visiting capitals, and meeting with heads of state; appeal for military aid; and showcase to its domestic constituents that the group had the backing of powerful allies abroad. Diplomacy was not determined by UNITA's fighting capacity; even as UNITA's armed strength increased with the help of U.S. assistance, UNITA maintained its diplomatic activism to ensure continued aid. When the end of the Cold War, abundant profits from diamonds, and defeat in the contest for international recognition eliminated further need for international or domestic support, rebel diplomacy correspondingly declined. The international community sought more than ever to use diplomatic means to negotiate a peace settlement with UNITA, but the latter responded with disinterest and belligerence. The rebels' political about-face after 1992 was matched by a tactical about-face in the war's diplomatic arena.

## Conclusion

Rebel diplomacy is a common feature of civil wars. While fighting against the government at home, rebel groups use a variety of means to conduct foreign affairs and rally overseas support. This article has shown that a political logic drives rebels' diplomatic activism. As actors seeking state power, rebel groups engage in diplomacy to demonstrate their ability to behave like states, adopt international norms, gestures, and rhetoric, and be seen in the world's diplomatic gatherings and capitals, with the end goal of gaining favor with important allies abroad. Diplomacy also boosts rebels' domestic legitimacy by assuring citizens that the group can attract international attention and aid and has the political and organizational tact required to implement overseas missions. Consistent with this logic, evidence offered in this article suggests that secessionist groups, for whom international recognition is essential for attaining independent statehood, and groups that organize domestically by investing in social service provision or creating legal political bodies, are more likely to become wartime diplomats. I found no empirical support for the notion that diplomacy is a weapon of the militarily weak or of the heavily repressed. Rather, it is employed by those groups that calculate that pursuing a two-pronged strategy of domestic and international legitimation offers the best chance of attaining their political objectives in the conflict.

Diplomacy, in short, is an important repertoire of contention for rebel groups. Armed confrontation is one dimension of a civil war; diplomatic confrontation is another, one in which states and rebels battle it out for external support. In an international system in which these issues matter practically for states' and rebels' very survival, the diplomatic arena can become a major site of contest between war opponents. Although scholarship has made major inroads toward understanding rebel groups' internal organization, relations with domestic populations, and strategic use of violence, little attention has been paid to their external strategies and relations with foreign audiences. This article shows that rebel groups' calculated engagement in international politics affects their wartime behavior. They internalize international relations in light of their own perceptions about how external forces interact with their political objectives. To neglect these international dimensions of rebellion is to miss a critical component of how nonstate entities conduct armed confrontations against their more formidable state opponents.

One fruitful extension of this research would be to examine the effects of rebel diplomacy. Does this rebelcraft make third-party intervention on behalf of the rebels more likely? Are groups that engage in it more likely to win wars than those that do not? What is the repertoire of state responses to rebel diplomacy? A promising avenue for future research is to explore these questions by addressing some of the empirical limitations of the present study. For instance, studies could examine the dynamic and interactive effects of foreign support on diplomacy by examining variation in these variables across time. Another important extension is to examine the dynamic aspects of rebel groups' political tactics in civil wars. Rebel diplomacy, certainly, is not new, but the opening of overseas offices may be a more recent tactic, one enabled by technological development. And if maintaining overseas offices is a twentieth-century invention in rebelcraft, then certainly rebels' use of websites, social media, and the blogosphere is a twenty-first-century phenomenon. Rebel groups have shown great sensitivity to changes in the international political, economic, social, and technological context, adapting their tactics accordingly. Beyond diplomacy, studying other rebel behavior in light of global developments should offer new insights into rebel groups' strategies of internationalization.

In addition to these theoretical insights, this article offers an important practical implication. If support-seeking is a strategy of warfare and it goads rebel groups to behave in ways that would grant them acceptance into the international community, then external actors wield significant leverage vis-à-vis rebel groups' wartime behavior. External aid, of course, has long been employed as an instrument of statecraft. Beyond this, states, international organizations, and nonstate actors have at their disposal a range of tools that affect rebel legitimacy, such as emphasizing the importance of compliance with international law or offering a rebel leader reception in a foreign capital as a carrot to motivate certain actions. This, in essence, is what diplomacy is all about. There is growing evidence that when states treat legitimacy-seeking rebel groups as if they are responsible states, they respond in kind.129 Consequently, creative thinking on the part of external actors in dealing with support-seeking rebel groups can have a significant impact on the latter's behavior, such as their treatment of civilians or their willingness to come to the negotiating table. In short, when targeted toward groups for whom international support can make or break their ability to attain their goals, wartime diplomacy can have payoffs in terms of human welfare and prospects for conflict resolution.

## Acknowledgments

For valuable comments and discussions, the author thanks Bridget Coggins, Rex Douglass, Jesse Driscoll, Tanisha Fazal, Carlos Fierros, Page Fortna, Hyeran Jo, Benjamin Jones, Morgan Kaplan, Douglas Lemke, Eleonora Mattiacci, Larry Napper, Idean Salehyan, Jack Snyder, Mohammad Tabaar, the anonymous reviewers, and participants at the various venues at which she presented earlier drafts of this article. Jacob Brahce, Breanna Irvin, and Emily Mullins provided excellent research assistance. Replication material and coding notes are available at http://dx.doi.org/10.7910/DVN/39DFOI.

## Notes

1.

“Syrian Opposition Will Have Foreign Mission in U.S.,” BBC News, May 5, 2014, http://www.bbc.com/news/world-us-canada-27287650.

2.

U.S. Department of State, “Daily Press Briefing” (Washington, D.C.: U.S. Department of State, May 5, 2014), http://www.state.gov/r/pa/prs/dpb/2014/05/225613.htm.

3.

Daniel Byman et al., Trends in Outside Support for Insurgent Movements (Santa Monica, Calif.: RAND Corporation, 2001), p. 47.

4.

Dushy Ranetunge, “London LTTE Chief Arrested,” Asian Tribune, June 22, 2007.

5.

Tommie Sue Montgomery, Revolution in El Salvador: From Civil Strife to Civil Peace (Boulder, Colo.: Westview, 1995), p. 114.

6.

Quil Lawrence, Invisible Nation: How the Kurds' Quest for Statehood Is Shaping Iraq and the Middle East (New York: Walker, 2008), p. 44.

7.

Denis M. Tull, “The Democratic Republic of Congo: Militarized Politics in a ‘Failed State,’” in Morten Boas and Kevin C. Dunn, eds., African Guerrillas: Raging against the Machine (Boulder, Colo.: Lynne Rienner, 2007), p. 120.

8.

Kwasi Nyamekye and Ralph R. Premdas, “Papua New Guinea–Indonesian Relations over Irian Jaya,” Asian Survey, Vol. 19, No. 10 (October 1979), pp. 927–945, at p. 943; and Peter Savage and Rose Martin, “The OPM in West Papua New Guinea: The Continuing Struggle against Indonesian Colonialism,” Journal of Contemporary Asia, Vol. 7, No. 2 (1977), pp. 338–346, at p. 338.

9.

George C. Herring, From Colony to Superpower: U.S. Foreign Relations since 1776 (Oxford: Oxford University Press, 2008).

10.

William C. Davis, The Secret History of Confederate Diplomacy Abroad (Lawrence: University Press of Kansas, 2005).

11.

On the difficulty and resource-intensity of transnational contention, see Victor Asal, Justin Conrad, and Peter White, “Going Abroad: Transnational Solicitation and Contention by Ethnopolitical Organizations,” International Organization, Vol. 68, No. 4 (September 2014), pp. 945–978.

12.

Patrick M. Regan, “Conditions of Successful Third-Party Intervention in Intrastate Conflicts,” Journal of Conflict Resolution, Vol. 40, No. 2 (June 1996), pp. 336–359; Dylan Balch-Lindsay, Andrew J. Enterline, and Kyle A. Joyce, “Third-Party Intervention and the Civil War Process,” Journal of Peace Research, Vol. 45, No. 3 (May 2008), pp. 345–363; David E. Cunningham, Kristian Skrede Gleditsch, and Idean Salehyan, “It Takes Two: A Dyadic Analysis of Civil War Duration and Outcome,” Journal of Conflict Resolution, Vol. 53, No. 4 (August 2009), pp. 570–597; David E. Cunningham, “Blocking Resolution: How External States Can Prolong Civil Wars,” Journal of Peace Research, Vol. 47, No. 2 (March 2010), pp. 115–127; Kristian Skrede Gleditsch, “Transnational Dimensions of Civil War,” Journal of Peace Research, Vol. 44, No. 3 (May 2007), pp. 293–309; and Idean Salehyan, Rebels without Borders: Transnational Insurgencies in World Politics (Ithaca, N.Y.: Cornell University Press, 2009). For an important exception, see Idean Salehyan, Kristian Skrede Gleditsch, and David E. Cunningham, “Explaining External Support for Insurgent Groups,” International Organization, Vol. 65, No. 4 (Fall 2011), pp. 709–744.

13.

In this I am joined by several recent works that specifically examine rebel diplomacy, including Bridget L. Coggins, “Rebel Diplomacy: Theorizing Violent Non-State Actors' Strategic Use of Talk,” in Ana Arjona, Nelson Kasfir, and Zachariah Mampilly, eds., Rebel Governance in Civil War (New York: Cambridge University Press, 2015), pp. 98–118; Morgan L. Kaplan, “Strategies of Insurgent Diplomacy: Evidence from the Middle East and North Africa,” University of Chicago, 2014; and Benjamin T. Jones and Eleonora Mattiacci, “A Manifesto, in 140 Characters or Fewer: Social Media as a Tool of Rebel Diplomacy in the Libyan Civil War,” paper presented at the International Studies Association annual conference, Toronto, Canada, 2014.

14.

See, for instance, Stathis N. Kalyvas, The Logic of Violence in Civil War (Cambridge: Cambridge University Press, 2006); Jeremy M. Weinstein, Inside Rebellion: The Politics of Insurgent Violence (New York: Cambridge University Press, 2007); Abdulkader H. Sinno, Organizations at War: In Afghanistan and Beyond (Ithaca, N.Y.: Cornell University Press, 2008); Zachariah Cherian Mampilly, Rebel Rulers: Insurgent Governance and Civilian Life during War (Ithaca, N.Y.: Cornell University Press, 2011); Paul Staniland, Networks of Rebellion: Explaining Insurgent Cohesion and Collapse (Ithaca, N.Y.: Cornell University Press, 2014); and Dipali Mukhopadhyay, Warlords, Strongman Governors, and the State in Afghanistan (New York: Cambridge University Press, 2014).

15.

Mampilly, Rebel Rulers; Reyko Huang, The Wartime Origins of Democratization: Civil War, Rebel Governance, and Political Regimes (Cambridge: Cambridge University Press, forthcoming); Clifford Bob, The Marketing of Rebellion: Insurgents, Media, and International Activism (Cambridge: Cambridge University Press, 2005); and Hyeran Jo, Compliant Rebels: Rebel Groups and International Law in World Politics (Cambridge: Cambridge University Press, 2015).

16.

For a full list of cases as well as values of the dependent and key independent variables used below, see the online appendix. The list of civil wars is adapted from Michael W. Doyle and Nicholas Sambanis, Making War and Building Peace: United Nations Peace Operations (Princeton, N.J.: Princeton University Press, 2006). Doyle and Sambanis use a relatively high threshold of deaths: 500–1,000 deaths in the first year, or at least 1,000 cumulative deaths in the first three years. For wars they coded as ongoing in 1999, I consulted the UCDP/PRIO Armed Conflict Dataset v.4–2010 to determine the wars' end dates. See Doyle and Sambanis, Making War and Building Peace; and Nils Petter Gleditsch et al., “Armed Conflict 1946–2001: A New Dataset,” Journal of Peace Research, Vol. 39, No. 5 (September 2002), pp. 615–637.

17.

See, for instance, the non-state actors data in Cunningham, Gleditsch, and Salehyan, “It Takes Two.”

18.

In most cases, the main group was easily identified (e.g., Fuerzas Armadas Revolucionarias de Colombia [FARC]; UNITA in Angola; and Front for the Liberation of the Enclave of Cabinda [FLEC] in Angola). In less clear-cut cases (e.g., the Algerian civil war of 1992–2002), the main group was identified by consulting secondary sources: in most of these wars, studies tend to converge on identifying one or two groups as being the dominant actors. Coding notes are contained in the supplementary material.

19.

Note, however, that a coding of 0 could be based on reliable evidence that the group did not, or could not, conduct diplomacy.

20.

Alex Vines, Renamo: Terrorism in Mozambique (Bloomington: Indiana University Press, 1991), p. 77.

21.

David Hoile, Mozambique: A Nation in Crisis (London: Claridge, 1989), p. 85; and Carrie L. Manning, The Politics of Peace in Mozambique: Post–Conflict Democratization, 1992–2000 (Westport, Conn.: Praeger, 2002), p. 93.

22.

Mohamed Kadamy, “Djibouti: Between War and Peace,” Review of African Political Economy, Vol. 23, No. 70 (December 1996), pp. 511–521, quotation at p. 250; and Peter J. Schraeder, “Ethnic Politics in Djibouti: From ‘Eye of the Hurricane’ to ‘Boiling Cauldron,’” African Affairs, April 1993, pp. 203–221, at p. 214.

23.

William Cyrus Reed, “International Politics and National Liberation: ZANU and the Politics of Contested Sovereignty in Zimbabwe,” African Studies Review, Vol. 36, No. 2 (September 1993), pp. 31–59, at p. 42.

24.

Fay Chung, Re-living the Second Chimurenga: Memories from the Liberation Struggle in Zimbabwe (Uppsala: Nordic Africa Institute, 2006), pp. 103–104.

25.

On the Maoists, see International Crisis Group, “Nepal's Maoists: Their Aims, Structure, and Strategy” (Kathmandu: International Crisis Group, 2005), p. 8; and Sudheer Sharma, “The Maoist Movement: An Evolutionary Perspective,” in Michael Hutt, ed., Himalayan People's War: Nepal's Maoist Rebellion (Bloomington: Indiana University Press, 2004), p. 40. On the Workers' Revolutionary Party, see Paul H. Lewis, Guerrillas and Generals: The “Dirty War” in Argentina (Westport, Conn.: Praeger, 2002), p. 45. On FRETILIN, see Geoffrey C. Gunn, Historical Dictionary of East Timor (Lanham, Md.: Scarecrow, 2011), p. 94; and John G. Taylor, East Timor: The Price of Freedom (New York: Zed, 1999), p. 77. On the Fedeyeen, see Helena Cobban, The Palestinian Liberation Organisation: People, Power, and Politics (Cambridge: Cambridge University Press, 1984). On the Dhofaris, see J.E. Peterson, Oman's Insurgencies: The Sultanate's Struggle for Supremacy (London: Saqi, 2007), p. 350.

26.

International Crisis Group, “Nepal's Maoists,” p. 8.

27.

Ezenwa-Ohaeto, Chinua Achebe: A Biography (Bloomington: Indiana University Press, 1997), p. 146; Roy M. Melbourne, “The American Response to the Nigerian Conflict, 1968,” Issue: A Journal of Opinion, Vol. 3, No. 2 (Summer 1973), pp. 33–42, at p. 34; and Bob, The Marketing of Rebellion, p. 26.

28.

Edward Aspinall, Islam and Nation: Separatist Rebellion in Aceh, Indonesia (Stanford, Calif.: Stanford University Press, 2009), p. 105; and Neil DeVotta, Blowback: Linguistic Nationalism, Institutional Decay, and Ethnic Conflict in Sri Lanka (Stanford, Calif.: Stanford University Press, 2004), p. 156.

29.

Paul Wallace, “Political Violence and Terrorism in India: The Crisis of Identity,” in Martha Crenshaw, ed., Terrorism in Context (University Park: Pennsylvania State University Press, 1995), pp. 352–409, at p. 395.

30.

David E. Cunningham, Kristian Skrede Gleditsch, and Idean Salehyan, “Non-State Actors in Civil Wars: A New Dataset,” Conflict Management and Peace Science, Vol. 30, No. 5 (November 2013), pp. 516–531.

31.

Mampilly, Rebel Rulers; and Tanisha M. Fazal, “Secessionism and Civilian Targeting,” paper presented at the annual meeting of the American Political Science Association, Chicago, Illinois, August 29–September 1, 2013.

32.

Bridget L. Coggins, Power Politics and State Formation in the Twentieth Century: The Dynamics of Recognition (New York: Cambridge University Press, 2014); and Tanisha M. Fazal and Ryan D. Griffiths, “Membership Has Its Privileges: The Changing Benefits of Statehood,” International Studies Review, Vol. 16, No. 1 (March 2014), pp. 79–106.

33.

Suresh C. Saxena, The Liberation War in Western Sahara (New Delhi: Vidya, 1981), p. 85.

34.

Nathan Gilbert Quimpo, “Back to War in Mindanao: The Weaknesses of a Power-Based Approach in Conflict Resolution,” Philippine Political Science Journal, Vol. 21, No. 44 (2000), pp. 99–126.

35.

Mampilly, Rebel Rulers, p. 247; Fazal, “Secessionism and Civilian Targeting”; and Jo, Compliant Rebels.

36.

Karen DeYoung, “Syrian Opposition Coalition Offices in U.S. Given ‘Foreign Mission’ Status,” Washington Post, May 4, 2014.

37.

Kalyvas, The Logic of Violence in Civil War; and Mampilly, Rebel Rulers.

38.

Mancur Olson, “Dictatorship, Democracy, and Development,” American Political Science Review, Vol. 87, No. 3 (September 1993), pp. 567–576.

39.

When they attempt to do so, by extorting money from states through kidnapping and ransom, for instance, they face further isolation and lose political credibility.

40.

On rebel diplomacy using Twitter, see Jones and Mattiacci, “A Manifesto, in 140 Characters or Fewer.”

41.

Awet Tewelde Weldemichael, Third World Colonialism and Strategies of Liberation: Eritrea and East Timor Compared (Cambridge: Cambridge University Press, 2012), pp. 207–209.

42.

“Syrian Opposition Leader Not Giving Up on Appeal for Anti-Aircraft Weapons,” Al-Monitor, May 9, 2014, http://www.al-monitor.com/pulse/originals/2014/05/syrian-opposition-anti-aircongress.html#.

43.

Bob, The Marketing of Rebellion.

44.

Fazal, “Secessionism and Civilian Targeting”; and Hyeran Jo and Catarina P. Thomson, “Legitimacy and Compliance with International Law: Access to Detainees in Civil Conflicts, 1991–2006,” British Journal of Political Science, Vol. 44, No. 2 (April 2014), pp. 323–355.

45.

Geraint Hughes, “A ‘Model Campaign’ Reappraised: The Counter-Insurgency War in Dhofar, Oman, 1965–1975,” Journal of Strategic Studies, Vol. 32, No. 2 (April 2009), pp. 271–305, at pp. 279–280.

46.

Aspinall, Islam and Nation, p. 239.

47.

Denis M. Tull, The Reconfiguration of Political Order in Africa: A Case Study of North Kivu (Hamburg: Institut für Afrika-Kunde, 2005), p. 123.

48.

Liz Sly and Karen DeYoung, “Libyan Rebels Get Diplomatic, Military Boost,” Washington Post, April 4, 2011.

49.

Tull, The Reconfiguration of Political Order in Africa, p. 286.

50.

Bridget L. Coggins, “Friends in High Places: International Politics and the Emergence of States from Secessionism,” International Organization, Vol. 65, No. 3 (July 2011), pp. 433–467; Coggins, Power Politics and State Formation in the Twentieth Century; Fazal, “Secessionism and Civilian Targeting”; Trace Lasley and Clayton Thyne, “Secession, Legitimacy, and the Use of Child Soldiers,” Conflict Management and Peace Science, Vol. 32, No. 3 (July 2014), pp. 289–308; and Jo, Compliant Rebels.

51.

Ibid.

52.

Fazal, “Secessionism and Civilian Targeting”; and Coggins, Power Politics and State Formation in the Twentieth Century.

53.

Fazal, “Secessionism and Civilian Targeting”; Lasley and Thyne, “Secession, Legitimacy, and the Use of Child Soldiers”; and Jo, Compliant Rebels.

54.

Mick Moore, “Thoroughly Modern Revolutionaries: The JVP in Sri Lanka,” Modern Asian Studies, Vol. 27, No. 3 (July 1993), pp. 593–642.

55.

Coggins, “Friends in High Places.”

56.

Fazal, “Secessionism and Civilian Targeting”; Jo, Compliant Rebels; and Reed, “International Politics and National Liberation.”

57.

Reed, “International Politics and National Liberation.” On empirical and juridical sovereignty, see Robert H. Jackson and Carl G. Rosberg, “Why Africa's Weak States Persist: The Empirical and the Juridical in Statehood,” World Politics, Vol. 35, No. 1 (October 1982), pp. 1–24.

58.

Mark Duffield and John Prendergast, Without Troops and Tanks: The Emergency Relief Desk and the Cross Border Operation into Eritrea and Tigray (Lawrenceville, N.J.: Red Sea, 1994), p. 28.

59.

Ibrahim Abdullah, “Bush Path to Destruction: The Origin and Character of the Revolutionary United Front/Sierra Leone,” Journal of Modern African Studies, Vol. 36, No. 2 (June 1998), pp. 203–235, at pp. 223, 224.

60.

Richard Chauvel, Nationalists, Soldiers, and Separatists: The Ambonese Islands from Colonialism to Revolt, 1880–1950 (Leiden, Netherlands: KITLV, 1990), pp. 376–378.

61.

Kirsten E. Schulze, “Insurgency and Counter-Insurgency: Strategy and the Aceh Conflict, October 1976–May 2004,” in Anthony Reid, ed., Verandah of Violence: The Background to the Aceh Problem (Singpapore; Seattle: Singapore University Press / University of Washington Press, 2006), pp. 225–271, at pp. 236–237.

62.

Salehyan, Gleditsch, and Cunningham, “Explaining External Support for Insurgent Groups.”

63.

Aspinall, Islam and Nation.

64.

Ibid.

65.

Colin Leys and John S. Saul, Namibia's Liberation Struggle: The Two-Edged Sword (London; Athens: J. Curry / Ohio University Press, 1995), p. 41.

66.

See Coalition website, http://en.etilaf.org/.

67.

Huang, The Wartime Origins of Democratization.

68.

Timothy P. Wickham-Crowley, “The Rise (and Sometimes Fall) of Guerrilla Governments in Latin America,” Sociological Forum, Vol. 2, No. 3 (Summer 1987), pp. 473–499; Weinstein, Inside Rebellion; and Mampilly, Rebel Rulers.

69.

Cunningham, Gleditsch, and Salehyan, “Non-State Actors in Civil Wars.”

70.

The NSA dataset contains a composite indicator for rebels' relative capability. Sixteen cases in my dataset, however, do not appear in the NSA dataset, and using the measure would thus lead to the exclusion of those cases from analysis. I therefore use one of the component measures of relative capability—rebels' mobilization capacity—and code the 16 missing cases with my own research on whether or not the rebels mobilized civilians (see coding notes for details). The other component measures—rebels' fighting capacity and ability to procure arms—would have been more difficult to code for the 16 cases, given ambiguity as to how these are defined and operationalized in the NSA data. Note also that the regression results reported in table 2, models 1 and 2 are robust to the exclusion of those 16 cases.

71.

Cunningham, Gleditsch, and Salehyan, “Non-State Actors in Civil Wars,” p. 522.

72.

Data on contraband are from Virginia Page Fortna, Does Peacekeeping Work? Shaping Belligerent Choices after Civil War (Princeton, N.J.: Princeton University Press, 2008). Data on rebel receipt of foreign material support are from the NSA dataset. For each variable, I supplemented the data with my own research and coding to fill in the missing observations. See coding notes for all documentation.

73.

Jennifer M. Hazen, What Rebels Want: Resources and Supply Networks in Wartime (Ithaca, N.Y.: Cornell University Press, 2013).

74.

Diplomacy might also be correlated with longer wars if it leads to greater external intervention. See Cunningham, “Blocking Resolution,” pp. 115–127.

75.

The 90 percent confidence interval is at 3 to 36 percent. The analysis was conducted using Clarify and is based on table 2, model 1. Discrete variables were set at their median and continuous variables at their mean, with K = 1000 simulations. The same setting applies to subsequent analyses of substantive effects. See Gary King, Michael Tomz, and Jason Wittenberg, “Making the Most of Statistical Analysis: Improving Interpretation and Presentation,” American Journal of Political Science, Vol. 44, No. 2 (April 2000), pp. 347–361.

76.

Ninety percent confidence interval at 10 to 42 percent.

77.

Ninety percent confidence interval at 1 to 40 percent. The joint inclusion of the rebel schools and legal political wings variables yields the same substantive results, with both variables statistically significant.

78.

The substantive effect is also fairly small: going from the 90th to the 10th percentile of the prewar Polity measure (i.e., going from a highly democratic state to a highly autocratic one) increases the likelihood of rebel diplomacy by only 13 percent; this is statistically significant at the 90 percent level.

79.

The results remain the same when the mobilization capacity variable and the two financing capability variables are separately included in the model.

80.

One potential bias arising from using a set of civil wars that excludes smaller conflicts, as I do here, is that because the dataset may be selecting on rebel strength, the effect of rebel capacity may be underestimated. There is, however, wide variation in rebel capacity within my set of cases.

81.

In a further analysis, I used receiver operator characteristics curves to assess the relative importance of the variables in terms of their in-sample predictive value. This analysis showed that although all of the included variables help explain rebel diplomacy, measures for rebel schools and secessionists carry the greatest explanatory weight. Full results are available from the author.

82.

On this method, see John Gerring, “What Is a Case Study and What Is It Good for?” American Political Science Review, Vol. 98, No. 2 (May 2004), pp. 341–354.

83.

For an in-depth account of this conflict, see W. Martin James III, A Political History of the Civil War in Angola, 1974–1990 (New Brunswick, N.J.: Transaction, 1992).

84.

At UNITA's inception, however, its policies were more Maoist. This might in part explain its vigorously anti-communist stance once it aligned itself with the United States.

85.

For instance, Jorge Sangumba served both as UNITA foreign minister and as its London representative in the 1970s, after which Tito Chingunji served in the two roles; Isaias Samakuva served as UNITA ambassador to Europe from 1998 to 1994 and again from 1998 to 2002; and Jeremias Chitunda served as the U.S. representative of the group from 1976 to 1986, after which Jardo Muekalia took over the role. See James, A Political History of the Civil War in Angola, 1974–1990; and Craig R. Whitney and Jill Jolliffe, “Ex-Allies Say Angola Rebels Torture and Slay Dissenters,” New York Times, March 11, 1989.

86.

Author's correspondence with Martin James, November 6, 2012.

87.

James, A Political History of the Civil War in Angola, 1974–1990, pp. 119, 177.

88.

“Apathetic Response to New UN UNITA Resolution,” Africa Analysis, November 14, 1997, p. 6.

89.

Library of Congress Name Authority File, http://id.loc.gov/authorities/names/n88043853.html. According to this source, the office began operations in 1988.

90.

U.S. House of Representatives, Angola's Government of National Unity, 41–594 CC, hearing before the Subcommittee on Africa of the Committee on International Relations, 105th Cong., 1st sess., April 24, 1997.

91.

“International Community Fails to Close UNITA Offices,” Angola: Peace Monitor, November 27, 1997, http://www.africa.upenn.edu/Urgent_Action/apic_12497.html.

92.

Prexy Nesbitt, “Terminators, Crusaders, and Gladiators: Western (Private and Public) Support for RENAMO and UNITA,” Review of African Political Economy, Vol. 15, No. 43 (1988), pp. 111–124, at p. 114.

93.

“International Community Fails to Close UNITA Offices.”

94.

James, A Political History of the Civil War in Angola, 1974–1990, p. 155.

95.

Ibid. See also “The Selling of Savimbi,” Africa Confidential, February 11, 1986.

96.

Letter, Jonas Savimbi to George Bush, March 2, 1986, OA/ID #19813-005, Donald P. Gregg Files Series, Bush Presidential Records, George H.W. Bush Presidential Library, College Station, Texas.

97.

Jonas Savimbi, “The Coming Winds of Democracy in Angola,” Heritage Lectures No. 217 (Washington, D.C.: Heritage Foundation, 1989).

98.

“Savimbi Says Bush Backs U.S. and Soviet Role in Angola Peacekeeping,” Associated Press, October 2, 1990.

99.

R.W. Apple, “Red Carpet for a Rebel, or How a Star Is Born,” New York Times, February 7, 1986. For a list of other lobbyists and PR firms that worked with UNITA, see Nesbitt, “Terminators, Crusaders, and Gladiators,” p. 122.

100.

Apple, “Red Carpet for a Rebel, or How a Star Is Born”; and W. Martin James III, Historial Dictionary of Angola (Lanham, Md.: Scarecrow, 2004), p. 135.

101.

White House, “United States Policy toward Angola,” National Security Decision Directive 274, May 7, 1987, Ronald Reagan Presidential Library, Simi Valley, California.

102.

Savimbi, “The Coming Winds of Democracy in Angola.”

103.

Jonas Savimbi, “Fighting for the Future of Angola,” Insight, January 20, 1986, pp. 74–79, at p. 79.

104.

See especially James, A Political History of the Civil War in Angola, 1974–1990. The fact that UNITA was fighting in a newly independent, autocratic, and highly unstable state likely helped lend credibility to its claims.

105.

Chester A. Crocker, High Noon in Southern Africa: Making Peace in a Rough Neighborhood (New York: W.W. Norton, 1992), pp. 172–173. U.S. Secretary of State Henry Kissinger had declared in 1976 that “there is every prospect of our dealing with the MPLA in Angola once it is clear that they are indeed an African government and not totally beholden of foreign input.” Quoted in Stefan Talmon, Recognition of Governments in International Law, Oxford Monographs in International Law (Oxford: Oxford University Press, 1998), p. 39.

106.

Ned Temko, “As Angolan War Grinds On, Foes Take Battle to Diplomatic Arena,” Christian Science Monitor, October 8, 1987, http://m.csmonitor.com/1987/1008/oango.html. See also James, A Political History of the Civil War in Angola, 1974–1990, p. 92.

107.

Savimbi, “Fighting for the Future of Angola,” p. 79. See also James, A Political History of the Civil War in Angola, 1974–1990, pp. 148–181.

108.

Peter W. Rodman, More Precious Than Peace: Fighting and Winning the Cold War in the Third World (New York: Charles Scribner's Sons, 1994), chap. 14; and James M. Scott, Deciding to Intervene: The Reagan Doctrine and American Foreign Policy (Durham, N.C.: Duke University Press, 1996), chap. 5.

109.

The NSA codes as “moderate” during this period UNITA's mobilization capacity, ability to procure arms, and fighting capacity relative to the government.

110.

Radek Sikorski, “The Mystique of Savimbi,” National Review, August 18, 1989, p. 35.

111.

See Jutta Bakonyi and Kirsti Stuvøy, “Violence and Social Order beyond the State: Somalia and Angola,” Review of African Political Economy, Vol. 32, Nos. 104–105 (June/September 2005), pp. 359–382, at p. 370; James, A Political History of the Civil War in Angola, 1974–1990, pp. 98–99, 119–120; Sikorski, “The Mystique of Savimbi,” p. 35; and John Marcum, “Government-in-Exile versus Government-in-Insurgency: The Case of Angola,” in Yossi Shain, ed., Governments-in-Exile in Contemporary World Politics (New York: Routledge, 1991), pp. 42–51, at p. 47.

112.

Bakonyi and Stuvøy, “Violence and Social Order beyond the State,” p. 370; and James, A Political History of the Civil War in Angola, 1974–1990, pp. 91–101.

113.

Victoria Brittain, Death of Dignity: Angola's Civil War (Trenton, N.J.: Africa World, 1998), p. 11.

114.

Gary van Staden, “Correspondent Describes Conference in Jamba,” Star, June 9, 1985, p. 10.

115.

David Zucchino, “Savimbi: A Revolutionary Who's Full of Contradictions,” Philadelphia Inquirer, July 1, 1988.

116.

Bakonyi and Stuvøy, “Violence and Social Order beyond the State,” pp. 370–371.

117.

Assis Malaquias, Rebels and Robbers: Violence in Post-Colonial Angola (Uppsala: Nordic Africa Institute, 2007), pp. 85, 109–110.

118.

Marcos Samondo, “Letter to the Editor: Don't Reward Angola with U.S. Recognition,” New York Times, February 10, 1993. Samondo was UNITA's representative to the UN.

119.

Malaquias, Rebels and Robbers, p. 104.

120.

Ibid.

121.

Quoted in Scott, Deciding to Intervene, p. 146. The hypothesis regarding authoritarian regimes is difficult to explicitly test through the UNITA case. Angola's Polity score at the start of the first conflict in 1975 stood at −7, but is coded as missing from 1991 to 1996 because of the political turmoil. The degree of repression likely did not change sufficiently to constitute a variable that could explain the differences in rebel diplomacy between the two conflict phases.

122.

Bakonyi and Stuvøy “Violence and Social Order beyond the State,” p. 372. These observations are consistent with Weinstein's argument that resource endowments shape rebel relations with civilians. See Weinstein, Inside Rebellion.

123.

Malaquias, Rebels and Robbers, p. 104.

124.

Ibid., p. 113.

125.

United Nations Security Council Resolution 1127, August 28, 1997.

126.

“Apathetic Response to New UN UNITA Resolution,” p. 6.

127.

Human Rights Watch, Angola Unravels: The Rise and Fall of the Lusaka Peace Process (New York: Human Rights Watch, 1999), p. 207.

128.

Ibid.

129.

Jo, Compliant Rebels.