Abstract

When rebel groups engage incumbent governments in war for control of the state, questions of international recognition arise. International recognition determines which combatants can draw on state assets, receive overt military aid, and borrow as sovereigns—all of which can have profound consequences for the military balance during civil war. How do third-party states and international organizations determine whom to treat as a state's official government during civil war? Data from the sixty-one center-seeking wars initiated from 1945 to 2014 indicate that military victory is not a prerequisite for recognition. Instead, states generally rely on a simple test: control of the capital city. Seizing the capital does not foreshadow military victory. Civil wars often continue for many years after rebels take control and receive recognition. While geopolitical and economic motives outweigh the capital control test in a small number of important cases, combatants appear to anticipate that holding the capital will be sufficient for recognition. This expectation generates perverse incentives. In effect, the international community rewards combatants for capturing or holding, by any means necessary, an area with high concentrations of critical infrastructure and civilians. In the majority of cases where rebels contest the capital, more than half of its infrastructure is damaged or the majority of civilians are displaced (or both), likely fueling long-term state weakness.

Introduction

The Libyan civil war began as a series of protests in the eastern city of Benghazi in early February 2011. Muammar Qaddafi's government lost control over most of the eastern province of Cyrenaica later that month. The opposition in the east formed the National Transitional Council (NTC) and announced leadership positions on March 5.1

On April 4, 2011, Italy announced its full recognition of the NTC as Libya's official government.2 This statement came only four days after the start of NATO's Operation Unified Protector, which had not yet dislodged Qaddafi's forces. Other states supporting the anti-Qaddafi rebels, including France and the United States, extended recognition to the NTC over the next three months.3 These bilateral recognitions carried material benefits for the NTC. Recognizing countries transferred frozen Libyan assets totaling billions of dollars to the new rebel government, all while the fighting on the ground continued.4 France and the United Kingdom also directly provided military aid to anti-Qaddafi forces, beyond the support they offered under Unified Protector.5 In exchange for recognition, the NTC's allies gained a geopolitical objective, the ouster of Qaddafi. Evidence also suggests that both the Italian and French governments used recognition of the NTC as part of a strategy to secure beneficial treatment of their oil companies after the war.6

Most states in the international system did not extend recognition to the NTC until late August 2011, however, when anti-Qaddafi forces began their assault on Tripoli.7 Between August 20 and a declaration of control over Tripoli on August 31, thirty countries recognized the NTC as Libya's new government.8 Despite their objections to NATO's role in Libyan regime change, Russia recognized the NTC on September 1 and China on September 12.9 On September 16, approximately two weeks after the fall of Tripoli, the NTC was given Libya's UN seat.10 By mid-September 2011, the NTC was thus treated as the sovereign authority of Libya within the most meaningful forum of international relations and by all major powers.11

At no point in the years since achieving international recognition has the NTC, or its successor, the General National Congress, possessed undisputed physical control over all Libyan territory.12 Although Qaddafi was killed on October 20, 2011, fighting has continued between the NTC and various armed groups at levels consistent with its designation as a civil war.13

The unfolding of events in Libya illustrates the significance of international recognition during civil conflict. States allied with anti-Qaddafi forces were able to use bilateral recognition of the NTC to affect the balance of capabilities in the conflict, both because they could transfer Libyan state assets to the rebels and because they could provide enhanced military assistance. Multilateral recognition by UN member states also generated both international legitimacy and material resources for the NTC while it struggled to establish control of the country after Qaddafi's demise.14

The Libyan case also highlights an underappreciated aspect of civil wars: political victory and military victory are not the same. Moreover, it is political victory that is the objective of combatants. Recognition establishes political victory in the international arena—it is a declaration of which individuals will be treated as the state's political authorities—and its timing is discretionary.15 Recognition is, thus, a powerful tool of foreign policy.

In the sixty-one center-seeking civil wars (i.e., sovereignty disputes over the entire state) initiated from 1945 to 2014, rebel groups achieved political victory (by securing bilateral recognition) prior to military victory in sixteen of the eighteen extra-constitutional transfers of sovereign authority.16 Nor were these new rebel governments on the brink of military victory when they achieved recognition. The median conflict continued 6.8 years following recognition by another state.17 Multilateral recognition at the United Nations—receiving a state's seat—follows a similar pattern. In thirteen out of the eighteen cases (including Libya), rebel challengers received UN seats before achieving military victory. This practice runs counter to claims by international legal scholars that recognition should be transferred only to rebel groups that establish “effective control,” defined as “undisputed paramountcy of physical power,” within their state.18

The motivating question of this article is: What explains the decision to accord one party or another status as the official government of a state in the midst of a civil war?

I evaluate two theories of recognition, both visible in the Libyan case: recognition driven by the prospects for economic and geopolitical gain; and recognition driven by a simple test, control of the capital city.

I process trace each of the sixty-one civil wars initiated from 1945 to 2014 to identify the moments when recognition occurs, as well as deliberate nonrecognition. I find that states generally have followed a decision rule conferring recognition on whoever holds the capital city.19 The spike in recognitions that the NTC received as rebel forces assaulted Tripoli is indicative of this broader pattern. States do, on occasion, prioritize their own geopolitical and economic gains over any standards of territorial control when deciding whom to recognize, as NTC allies did, but this behavior is rare, and most cases involve the United States withholding recognition from communist governments during the Cold War.

The general decision rule to accord recognition based on capital city control has profound consequences during civil war. In particular, it generates rational incentives for combatants to use any means necessary to gain or maintain hold of the capital's underlying territory. There is, however, no incentive to keep the city intact. I find that in twelve of the nineteen cases in which rebels succeeded in contesting the capital, more than 50 percent of the city's infrastructure was destroyed or the majority of the population was displaced, or both. Mogadishu, Brazzaville, and Kabul are all examples of capital cities reduced to rubble in the service of achieving international recognition.20 Evidence suggests that combatants are aware of this incentive structure surrounding recognition. For example, William Reno documents how Charles Taylor fixated on capturing Monrovia as the key to becoming “the internationally recognized president,” despite establishing a significant personal fiefdom within the bulk of Liberian territory.21 Repeated assaults on Monrovia over the course of the war damaged all of its major infrastructure.22

This targeting of the capital city also has long-term consequences for state strength. First, the destruction of the city and the displacement of the population weakens the state itself, given that the basic infrastructure of these cities is a prerequisite for the delivery of public goods, both locally and statewide.23 Second, setting the bar for political victory low—at territorial control of the capital—may also lower the perceived costs of initiating conflict to gain political authority in the first place, thus weakening the strength of electoral institutions.

Given the perverse incentives created by the capital control test, this article also considers briefly the historical precedents for altering the standards of international recognition. Based on these precedents, I suggest that states collectively can alter combatant incentives if they are willing to make two changes to current recognition practice: removing territorial control as a sufficient condition for recognition and formalizing the use of nonrecognition at the UN. These two changes allow new recognition criteria, such as a necessary minimal standard of battlefield conduct, to have teeth.

This article contributes to the limited literature on recognition of governments, arguing that it should be studied as a meaningful tool of foreign policy. It also identifies a pattern in recognition behavior during civil wars with consequences for the conduct of conflict. Finally, it provides a starting point from which to debate the standards of international recognition, by arguing that the current procedure is driven by convenience, not necessity, and that alternatives are plausible.

The article proceeds in seven sections. First, I review the foundational scholarship on the recognition of governments within international law, to provide definitions of the core concepts. Second, I examine the existing literature regarding the significance of recognition in the study of civil war. In the third section, I introduce the two models of recognition behavior. The fourth section discusses the data, including case identification and variable coding. In the fifth section, I review civil wars in three countries—Angola, Cambodia, and Chad—where primary and secondary sources provide insight into the drivers of bilateral and multilateral recognition. The sixth section tests whether the conclusions drawn from these case studies travel to the remainder of the data. The final section concludes with a discussion of policy implications.

Recognition in International Law

M.J. Peterson describes recognition as “a discretionary act producing binding effects,” where these “binding effects” are manifested as consequences in law.24 There are two kinds of recognition within international politics: recognition of states and recognition of governments. Although recognition of one is not entirely independent of the other, they are conceptually distinct.25 I exclude secessionist civil wars (and, thus, the creation of new states) from this analysis because they contain both types of recognition, which are hard to disentangle within a single case. I do discuss the state recognition literature in the next section, however, as it bears on the recognition of governments.

According to the Foreign Affairs Law of the United States, recognizing a government is a “formal acknowledgment that a particular regime is the effective government of a state and implies a commitment to treat that regime as the government of that state.”26 In this formulation, “effective” does not mean “functionally true.” Rather, it means an entity with legal standing.27

The Foreign Affairs Law stipulates that “treating a regime as a government includes accepting its acts as creating international rights and obligations.”28 In practice, governments have a certain bundle of privileges that accrue through this recognized sovereign authority. These include possession of the legitimate means of violence, entitlement to the disbursement of state assets, legal standing in international courts, the ability to create binding contracts on behalf of the state, and the right to designate the state's representatives in international institutions.29 In short, recognition of a government by another state refers to the acknowledgment that a particular set of individuals has the unique legal authority to commit a number of significant acts on behalf of the state. Throughout this article, I use the terms “recognition” or “full recognition” to refer to this formal, legalistic act.

When a state undergoes a constitutional change of government, recognition is not generally a subject of discussion. When the change occurs outside the bounds of a state's constitutional process—through conquest, popular revolution, or coup, for example—recognition comes into question.30 In this context, when states recognize a new government before it has achieved effective control over the country's territory, it is described as “premature” and a violation of international law.31 The “premature” label also applies to recognition of a new rebel government while the incumbent is still putting up a fight in state territory.32 Recognition—particularly if it is premature—is fundamentally a constitutive act for weak governments.33 When other states deem a particular government to be the bearer of the sovereign rights and privileges, that government then holds those rights independent of its capacity to act on them.34

Recognition can occur on several levels of interstate relations. Bilateral recognition creates a legal entity for the purpose of the recognizer's domestic courts. This entity can initiate lawsuits, possesses immunity from lawsuits in some contexts, is generally entitled to state property within the recognizer's jurisdiction, and is able to enter into bilateral agreements.35

Multilateral forums represent a separate context in which states can accord recognition to one another.36 For example, representatives of United Nations member states—by virtue of accepting another representative's credentials—collectively determine who has the right to speak on behalf of that member state.37 UN-level recognition is treated as a signal of membership in the international community of governments as a whole.38 The most significant legal consequence of multilateral recognition at the UN level is the access to funds and assistance provided through UN organizations and affiliates such as the World Bank, which deal only with sovereigns.

Legal scholars note that there is no inherent requirement for states to extend recognition to another government bilaterally or in a multilateral forum, even the most territorially established one.39 Thus, in either setting recognition is discretionary.

During civil war, recognition has a number of consequential legal effects. The language states use to avoid granting full recognition to rebels indicates this is well understood.40 First, it legitimizes certain military entities and delegitimizes others, including third-party interveners.41 Second, recognition determines which combatants are entitled to draw on state assets to aid their cause.42 In practice, assets controlled by foreign governments and corporations can be substantial and their disposition entirely determined by recognition. For these reasons, bilateral premature recognition itself is classified, in international legal terms, as an act of intervention in civil war.43 Multilateral recognition also permits embattled regimes to request intervention forces and financial support.44 Both bilateral and multilateral recognition, therefore, have legal consequences that can affect the course of civil war.

Recognition in Civil War Scholarship

Existing scholarship on recognition has focused, more often than not, on the issue of recognizing new states.45 Although new state recognition is conceptually distinct, the standards are suggestive for the recognition of new governments. Specifically, there is a similar tension between according recognition based on shared decision rules and according recognition based on political expediency.

The most comparable situations to the recognition of governments during civil war are those arising from wars of secession.46 In the period from 1815 to 1945, James Crawford argues that facts on the ground, in terms of independence and effective territorial control, played a decisive role in third-party recognition.47 Effectiveness as a criterion for new state recognition did not appear especially relevant after 1945, however, particularly with the new states arising from decolonization.48 Crawford argues that in the modern era, when the entity is not a “self-determination unit” or colony, recognition of a new state out of political expediency does not occur unless the home state grants permission.49 This suggests that the motivations of third-party recognizers may be checked by certain tacit rules.

Bridget Coggins, however, argues that the political motivations of third-party recognizers can be decisive when the creation of new states is viewed as a social process.50 Coggins also highlights that this politicized recognition of new states can have consequences at the level of the international system, because more states begin life internally weak than would be the case if effectiveness were a meaningful criterion.51

Scholarship on the recognition of governments has primarily focused on the reasons why rebel groups pursue recognition.52 One motive is economic. As Jennifer Hazen argues, access to resources is a critical, dynamic element of rebel group capacity during civil war.53 Incumbents have privileged access to funds, resulting in part from their sovereign status. For example, multinational firms engaged in monetizing natural resources prefer to deal with sovereigns as a risk-mitigation strategy.54 Long-term extraction of resources requires contracts that can be enforced, and enforcement requires that both parties to the contract have legal standing. Although preemptive deals for domestic natural resource exploitation are occasionally struck with rebel groups, these agreements are relatively rare.55 Other economic advantages of sovereignty during civil war include access to international capital markets, specifically for large-scale borrowing.56

A second reason rebel groups pursue recognition concerns the military balance. Odd Arne Westad has noted that recognition can enable a shift from receiving covert to overt military aid, and that the scale of overt aid can be substantially greater.57 Reyko Huang has shown more generally that some rebel groups devote considerable resources toward securing benefits from third parties, including but not limited to recognition, as a tactic within their overall campaign.58

This existing literature thus sheds light on the motives for combatants seeking recognition during civil war. The motives of third-party recognizers, however, remain an open question.

Models of International Recognition

In this section, I outline two models of government recognition during civil war that echo the debate surrounding the recognition of new states: a model of self-interest and a model of capital city control. Each model makes different predictions about the conditions under which states will extend recognition to rebel governments during civil wars.

A MODEL OF SELF-INTEREST

As stated at the outset, recognition is a powerful tool of foreign policy. With this tool, states wield the power to ratify and even construct their peer governments.59 It is plausible, then, that states use recognition to advance their interests within the international system and to shape the membership of the system itself.

As with the recognition of new states, self-interest may derive from the geopolitical advantages gained by recognizing friendly governments or withholding recognition from unfriendly ones. Alternatively, as Reno suggests, self-interest also can take the form of economic gains, generally through preferential access to resource exploitation.60 The Libya case provides an example of both geopolitical and economic benefits. France and Italy stood to gain economically by early recognition of the NTC through their national oil companies. Other NTC allies, the United States included, perceived a geopolitical benefit from replacing Qaddafi with a friendlier alternative.61

A self-interested actor may still be concerned about getting recognition wrong, however. On the one hand, nonrecognition of an eventual victor can alienate the new government.62 On the other hand, extending recognition to an entity that is ultimately defeated on the battlefield can be embarrassing.63 Therefore, the decision to recognize requires weighing risks against rewards.

If this model accurately captures the motivations behind state recognition practice, one is most likely to see premature recognition—recognition at its most discretionary—when gains are available and the risk of backing the wrong actor is minimal.

THE CAPITAL CITY CONTROL TEST

The self-interest model does not ascribe significance to a decision-rule approach to recognition, particularly in the presence of net benefits to the recognizer. As Coggins points out with respect to the creation of new states, however, there are reasons to suspect that the arbitrariness afforded by a model of self-interest would pose a significant problem for the current state system.64

I propose that, since 1945, states have relied on a model of recognition with a minimal decision rule: control of the capital city as a necessary and sufficient condition for international recognition.65

There are several reasons why I argue that recognition may be a rule-driven sphere of international relations in general and why I hypothesize that the capital control test specifically guides recognition practice. First, rule-based recognition appears to have been the norm prior to World War II, though what constituted the “rules” evolved over time.66 Second, there appears to be a peacetime rule in place for sovereign authorities in postcolonial states, where control of the capital is sufficient for continued recognition of a government.67 This minimal standard is justified in peacetime partly by the idea that control over the empirical state—including the army, the bureaucracy, the treasury and central bank, and major infrastructure such as airports and radio stations—only requires holding the capital, though in practice this is often not true.68

The capital city control test is also well suited to the post–World War II international environment where there is “demand” for continuous representation for every state in the international system. Multinational organizations such as the United Nations require daily contact between a set of individuals who are generally agreed upon to represent their states.69 At the same time, the supply of ambiguous representation cases has increased with the number of states and civil conflicts. A minimal decision rule makes the conduct of international relations easier under conditions of uncertainty. Specifically, the capital control test reduces the problem of identifying legitimate sovereigns down to a single, observable dimension, much like a single-variable heuristic.70

If the capital control model accurately captures state recognition practice, then one should see recognition of new governments occur only once the capital has changed hands, even if third parties have an economic or geopolitical incentive to confer recognition sooner.

Methodology and Data

I weigh the explanatory power of the self-interest and capital city control models by identifying and process tracing the recognition events associated with all center-seeking civil wars initiated from 1945 to 2014. I draw the data for these cases from primary documents, secondary sources, and journalistic accounts. This section defines the universe of cases, the typology of actors used across cases, and the operationalization of the variables.

IDENTIFICATION

Identification of political recognition and its relationship to the military progress of a conflict requires a particular understanding of civil war. In the literature, the definition of civil war, as well as measurements of its onset, duration, and termination, are somewhat contested.71 My analysis relies on two restrictions around the definition of civil war that are consistent with some, but not all, existing perspectives.

First, I take center-seeking civil wars to be sovereignty disputes over the entire state.72 As such, I measure onset, duration, and termination of a given war at the state level. I consider only violence that is center seeking. That is, the objective of the violence is gaining or retaining control of the entire state. This definition excludes from consideration any violence associated with secessionist objectives. In addition, this approach stands in contrast to conflict measurement along government-rebel group dyads and puts my definition (and number of cases) at odds with some recent analyses.73

Second, I isolate the military dimension of the conflict to define when a state is experiencing a civil war. Doing so is consistent with literature focusing on battle deaths, but not with classification systems that restart civil wars upon shifts in sovereign authority.74 The reason for the focus on military outcomes as defining the period of a war is the nature of my research question. The timing of recognition—a political development—is measured relative to the military progress of the conflict.

Given these two restrictions, I begin with James Fearon's dataset of civil war durations, which is territorially structured (rather than dyadic) and has relatively high battle-death onset and duration criteria.75 I exclude coups, popular revolutions, secessionist wars, and certain cases of indeterminate coding.76 I augment this dataset with civil wars beginning after 1999, as well as battle-death-based end-dates, drawn from the UCDP Battle-Related Deaths Dataset v.5-2015 and the UCDP/PRIO Armed Conflict Dataset v. 4-2015, 1946–2015.77 Aggregating the relevant cases results in a universe of sixty-one center-seeking civil wars with onset from 1945 to 2014. Some cases represent consolidation of wars where battle deaths remain high over time, but which have been semantically separated in the literature (e.g., the First and Second Congo Wars).78 I also include a war in Somalia after the recognized government disappears in 1991, as this fighting is still very much consistent with a contest for sovereign authority of the state.79 The list of wars is provided in appendix A.80 Unless otherwise stated, I refer to all conflicts as they are named in appendix A.

For each of these sixty-one cases, I use primary and secondary sources to process trace the retention or transfer of sovereign authority over the course of the conflict.81 Seventy-five percent of center-seeking civil wars in this period do not experience any extra-constitutional transfer of sovereign authority, and no third-party state makes a claim that such a transfer should occur. In these cases, the incumbent retains power or the parties eventually reach a settlement. Two wars (Afghanistan and Chad) contain more than one transfer. The recognition case list and coding is provided in appendix B.82

ACTORS: ALLIES, OPPONENTS, AND NEUTRALS

To create a consistent typology of actors across cases, I draw on the belligerency literature.83 I define three categories of actors: allies, opponents, and neutrals, relative to a rebel group. Allies are those states that provide military or nonmilitary aid to the rebel group. Opponents are those states that provide military or nonmilitary aid to the incumbent government. These two types of actors perceive there to be some benefit from their preferred party gaining or retaining recognition. Neutrals are the residual category containing all other states that, by definition, do not participate in the conflict. I identify allies and opponents using the primary and secondary sources for each war of interest. In some cases, no state satisfies ally or opponent criteria.84

Allies have—at minimum—an interest in making geopolitical gains from their preferred party toppling the incumbent. In some cases, they may have the added benefit of economic gains. As supporters of the incumbent, opponents risk a net loss at least in geopolitical terms, but possibly in economic terms as well, by recognizing a rebel government. Premature recognition serves only to accelerate that loss.

Neutral states may still reap benefits of regime change (or bear costs). Given that their interests are indeterminate as a group, I focus primarily on the behavior of allies and opponents.

DEPENDENT VARIABLES

In the post–World War II era, states have two opportunities to accord recognition. First, they can choose to bilaterally recognize another government. Second, they can accept or protest the credentials of a state's representative in an international body such as the United Nations. Legal scholars consider these to be independent acts, so one does not necessarily affect the other in law.85 To understand recognition practice as a whole, I examine both of these outcomes separately.

bilateral recognition. Bilateral recognition is measured through public statements of recognition, including press releases and speeches, as well as forms of diplomatic contact that constitute recognition.86 Statements providing less than full recognition are excluded, because they do not carry legal weight.87 I focus on precise (dated) measures of recognition for allies and opponents, but also report neutral state data where available. In addition, I note those moments when a large number of countries recognize a new government.88

multilateral recognition. Multilateral recognition is measured by credentials acceptance, protest, or blocking at the United Nations.89 As there is no official rule in the UN Charter to determine which governments are judged to have the right to seat representatives for a given state, it is an inherently flexible forum for recognition.90

In practice, the UN Credentials Committee is responsible for evaluating the credentials of each state's representative and declaring them in order. These credentials must be “issued either by the Head of the State or Government or by the Minister of Foreign Affairs.”91 The power of the Credentials Committee is evidenced by the practice of states that circumvent formal rules to remain on it. The United States and Russia (and, previously, the Soviet Union) have sat on all but one Credentials Committee since 1950, though no rule guarantees that arrangement and it contravenes the practice of rotating memberships.92 China has also achieved de facto permanent membership on the Committee.93

The nine-member Committee produces a report at the beginning of each UN session recommending which credentials should be accepted (or rejected) by the General Assembly. Within the Committee, the motion to put forward the report is based on a majoritarian vote where no member has a veto. Although Committee members cannot unilaterally block the acceptance of credentials, they can and do go on the record to protest cases with which they do not agree.94 The General Assembly then votes on a resolution to accept the Committee's recommendations. Although the General Assembly has never voted not to accept the report of the Credentials Committee, individual states have motioned to do so or registered their protests.95

Credentialing at the UN thus provides a way of observing third-party state behavior with respect to multilateral recognition. Committee members have the power to propose rejecting a contested set of credentials. Within the General Assembly, other states can register their protest of a representative's credentials during the vote to accept the Credentials Committee's report.96 Using the annual Reports of the Credentials Committee, I record whether or not a particular case of credentials was debated and the vote for and against acceptance. I also record whether or not there were any objections to adopting the Report of the Credentials Committee in the General Assembly for each relevant session. As the Committee approves credentials at the start of each annual United Nations session, usually between October and December, I consider every meeting of the Committee (and subsequent General Assembly vote) while a civil war is ongoing to be a potential observation of multilateral recognition.97 In rare cases, Security Council members have the opportunity to block a government from UN representation when the issue involves simultaneous recognition of a new state, given that a state's application to join the UN must pass through the Security Council. This situation arises twice in the data, with respect to Angola and Tajikistan.98

A caveat with regard to nonfindings is required. UN documents tend to record the topics of discussion in the context of voting on member credentials, in either the Committee or the General Assembly. Therefore, in the case of multilateral recognition, if I find no discussion of recognition in the documents, I can be reasonably confident recognition was not controversial in that case. For bilateral recognition, however, I cannot rule out the possibility that an observation of nonrecognition is the result of incomplete research, rather than an absence of fact.

INDEPENDENT VARIABLES

In process tracing these conflicts, I considered three potential drivers of recognition: total control over territory and population; control of the capital city; and expected gains or losses for ally and opponent states.

I evaluated the degree of overall territorial control established by rebel groups over the course of the conflict using both primary and secondary sources.99 I code capital city control as taking all physical territory within the internationally recognized capital, including major landmarks (e.g., the presidential palace).100 Incumbents have a built-in advantage, because they can lose a portion of the capital without losing overall control. Both short- and long-term cases of partial control are coded as being incomplete.101

Allies and opponents, by definition, have geopolitical interests in the conflict. I consider the availability of economic gains (or the potential for losses) as a separate kind of incentive for recognition (or denial thereof). I operationalize economic gains as the presence of natural resources available for exploitation by allies. Twenty-eight wars occurred in states with some resource extraction industry or known petroleum reserves.102 Using primary and secondary sources, I code whether rebels gained control over the means to exploit those resources over the course of the conflict.103 I consider economic gains to be available only if resources are present and under some degree of rebel control.

Case Studies

During a conflict, recognition unfolds over time. I use civil wars in three countries to illustrate this process and the specific claims states make to justify their recognition choices. I selected cases to balance the availability of a rich documentary record with sound comparisons. The simultaneous declaration of two governments of Angola upon independence in 1975 provides unique insights into bilateral recognition conditions. A comparison of Angola with the contemporaneous first Cambodian civil war (1967–75) allows for some control over key policymakers and the geopolitical environment, and offers a clearer picture of the multilateral recognition process, which occurred over three sessions at the UN. Finally, the first two Chadian civil wars demonstrate that variation in outcomes is not solely the result of variation in the country of interest.

ANGOLA, 1975–2003

Although recognition of the first government of independent Angola is inextricably tied to its recognition as a newly independent state, the issue of government recognition, particularly at the bilateral level, can be considered separately, because there appears to have been no question in the international community that Angola should become an independent state.104 The Angolan case provides a test not only of the relative weight of self-interest and control of the capital city, but also of the commitment to evaluating a government based on its effective control over the totality of the state's territory.

timeline. In January 1975, the Portuguese government committed to granting Angola its independence on November 11, 1975.105 The three major militarized independence movements nominally became participants in a transition government prior to independence, but in practice fought one another for an absolute upper hand. The Movimento Popular de Libertação de Angola (MPLA), led by Agostinho Neto, and the Frente Nacional de Libertação de Angola (FNLA), led by Holden Roberto, both had forces in the internationally accepted capital, Luanda. Jonas Savimbi's União Nacional para a Independěncia Total de Angola (UNITA) was based in the countryside.106

The MPLA was avowedly communist and received overt support from Cuba, including weapons and troops, and covert aid from the Soviet Union.107 The United States initially supported the FNLA with covert monetary aid, but also began supporting UNITA in July 1975.108 China supported the FNLA, and a number of regional powers in Africa provided military and logistical support to all three movements.109

On June 27, 1975, the issue of which group would be victorious was raised at a meeting of the National Security Council.110 Director of Central Intelligence William Colby presented a paper arguing that “military control of Luanda by either group would not necessarily determine control of or influence over the rest of Angola, particularly in the rich agricultural areas or along all the main transportation routes.”111 When President Gerald Ford asked Secretary of State Henry Kissinger his view, however, Kissinger replied:

I do question the judgment that control of the capital is not of importance. The history of Africa has shown that a nation's only focal point is the capital, and whoever has the capital has a claim on international support. In the Congo civil war, the reason we came out on top is because we never lost Leopoldville. If Neto can get Luanda, and drive the others out, he will have a power base, and gradually gain support of other Africans.112

Neto and the MPLA dislodged the FNLA from Luanda in mid-July 1975.113 But by early November, the MPLA controlled only Luanda and a belt of territory from the coast to Zaire, while the FNLA held positions to the north, and allied FNLA and UNITA forces held the southern half of the country.114

Despite his statement to the president in June, days before the Portuguese withdrawal in November, Kissinger cabled U.S. embassies worldwide asking them to press an “effective control” argument with their counterparts:

WE WOULD VIEW A CLAIM BY THE MPLA THAT IT IS THE GOVERNMENT OF ANGOLA, AND THAT IT MERITS RECOGNITION AS SUCH, AS UNFOUNDED, BECAUSE IN FACT IT CONTROLS A MINORITY OF THE POPULATION AND TERRITORY OF ANGOLA AND CONSEQUENTLY IS NOT IN A POSITION TO REPRESENT THE INTERESTS OF THE STATE OF ANGOLA INTERNATIONALLY. RECOGNITION BY THE INTERNATIONAL COMMUNITY OF THE MPLA REGIME AS THE ANGOLAN GOVERNMENT COULD ONLY SERVE AS FURTHER OBSTACLE TO SEARCH FOR A PEACEFUL, NEGOTIATED SOLUTION AND TO VALIDATION OF PRINCIPLE OF SELF-DETERMINATION. THE CLAIM OF FNLA/UNITA WOULD, IN OUR VIEW, DESERVE TO BE TAKEN INTO ACCOUNT BY THE INTERNATIONAL COMMUNITY IN WEIGHING POSSIBILITIES AND TIMING OF RECOGNITION OF THE STATE OF ANGOLA AND OF THE GOVERNMENT WHICH ACTUALLY EXERCISES EFFECTIVE CONTROL OF THAT STATE.115

The replies sent from U.S. embassies provide some indication of the relative persuasiveness of the “effective control” argument, in contrast to Kissinger's own position in June that the capital would be decisive.

From the embassy in Khartoum:

IN RESPONSE MY QUESTION, FONMIN [foreign minister of Sudan] STATED THAT GOS [the government of Sudan], LIKE USG [the U.S. government], PLANS REFRAIN FROM MAKING ANY HASTY MOVE TOWARD RECOGNIZING MPLA … HE FEELS NUMBER AFRICAN STATES MAY RECOGNIZE MPLA SIMPLY BECAUSE IT HOLDS THE CAPITAL, IN HOPE THAT GROUP'S CURRENT RADICALISM WILL BE TEMPERED BY RESPONSIBILITY.116

The embassy in Paris equated holding or gaining the capital with victory in the civil war:

THE QUAI [the French ministry of foreign affairs] ESTIMATES THAT THE MPLA NOW CONTROLS LESS THAN A THIRD OF THE TERRITORY AND POPULATION OF ANGOLA. GIVEN LUANDA'S PSYCHOLOGICAL IMPORTANCE NETO WILL PUT ALL HIS EFFORT INTO HOLDING IT AND THE FNLA/UNITA ARE NOT LIKELY SOON TO ACHIEVE VICTORY.117

From the embassy in Tanzania:

INFORMATION: THOUGH TANGOV [the government of Tanzania] AND MEDIA CREDIT MPLA CLAIMS TO CONTROL MOST OF ANGOLA, TANGOV ALREADY FULLY AWARE OF EVIDENCE TO CONTRARY … COMMENT: TANGOV WILL BE AMONG THE FIRST TO RECOGNIZE MPLA AS GOVERNMENT OF ANGOLA.118

U.S. allies provided some acknowledgment of “effective control,” in theory. From the embassy in Canada:

GOC [the government of Canada] IS NOT RECOGNIZING ANY GOVERNMENT PENDING THE ABILITY OF A GOVERNMENT TO DEMONSTRATE THAT IT HAS EFFECTIVE CONTROL OVER THE TERRITORY AND THAT IT IS ABLE TO HONOR ANGOLA'S INTERNATIONAL OBLIGATIONS.”119

Both the MPLA and a UNITA/FNLA coalition announced new governments of Angola on November 11, 1975. The MPLA announced its capital as Luanda, while the coalition announced its capital as Huambo.120 Within two days, a dozen states had extended bilateral recognition to the MPLA as the official government of Angola, and more allied and neutral states granted recognition through the end of the year.121 No state, including the United States or other allies providing military aid, extended bilateral recognition to the UNITA/FNLA coalition.

Passage of the Tunney amendment in December 1975 restricted Kissinger's ability to fund UNITA, but he continued the diplomatic pressure not to extend recognition to the MPLA.122 Kissinger also put pressure on the American firm Gulf Oil to stop payments due to the government of Angola from their operations in the oil enclave of Cabinda, hoping to weaken the MPLA.123 France promised to continue military aid to UNITA, including Mirage helicopters and French pilots.124 Neither France nor the United States extended formal recognition, however.

In early February 1976, UNITA lost control of its provisional capital Huambo.125 The Organization of African Unity voted on February 10, twenty-seven to nineteen in favor of admitting Angola as represented by the MPLA government.126 Recognition by the United States' major allies, including Britain, Canada, and France, soon followed, though UNITA maintained a commitment to resistance and the MPLA still did not control all Angolan territory.127 In justifying its new position to the United States, France cited the desire to support its regional allies that were also planning to recognize the MPLA.128

The United States withheld bilateral recognition of the Angolan government and, in June 1976, vetoed Angola's entry into the United Nations, thereby denying the MPLA government multilateral recognition, with an understanding this would irritate U.S. allies.129 Six months later, however, the United States abstained from a second vote regarding Angolan membership, and the MPLA-led government gained representation at the United Nations.130

The United States never subsequently challenged the credentials of the MPLA's representative. Yet, it did not extend bilateral recognition to the Angolan government until 1993.131 The civil war itself did not end until 2002, after the death of Jonas Savimbi and negotiation of a settlement.132

analysis. Bilateral recognition of the MPLA by its allies is consistent with both the self-interest and capital city control models of recognition. The Soviet Union and Cuba gained an allied government by recognizing the MPLA, and their preferred group also held the official capital. Their interest was largely geopolitical; an American firm benefited most from Angola's oil resources at the time, and no move was made to displace Gulf Oil. Still, neither majority control over the territory and population nor battlefield momentum proved sufficient grounds for UNITA's allies to extend recognition, suggesting pure geopolitical interest was not enough to justify that action. As an MPLA opponent, however, the United States appears to have prioritized geopolitical gains by delaying recognition of the MPLA at the UN (and denying it bilaterally for eighteen years). Other states, including Canada, France, and UNITA's regional allies, appeared satisfied with the MPLA's less-than-total accomplishments on the battlefield, though it was known that Soviet and Cuban assistance had made these achievements possible.133

The Angolan case thus suggests that holding the capital may be necessary for positive recognition by allies, though it may not always be sufficient with respect to opponents.

CAMBODIA I, 1967–75

In June 1975, Secretary of State Kissinger argued that the significance of the capital city for bilateral recognition was a feature only of African politics.134 The first Cambodian civil war between Lon Nol's Khmer Republic government and the Khmer Rouge suggests otherwise. Given that the cases included recognition events less than one year apart, many of the individuals involved (including Kissinger) and the prevailing geopolitical conditions were the same. The case also provides a clearer look at the arguments around multilateral recognition and its timing.135 As with Angola, economic benefits of recognition for allies were not a significant feature of the conflict, but geopolitical concerns were. The progression of rebel control from the periphery to the capital also illustrates how the city can be treated as a necessary condition of recognition, both bilateral and multilateral.

timeline. The war began as a rebellion against Prince Sihanouk's government, primarily in Cambodia's periphery in 1967. In March, 1970, Prime Minister Lon Nol and the National Assembly orchestrated a coup while Sihanouk was in Moscow. Lon Nol's post-coup government received Cambodia's UN seat without comment.136

Sihanouk's government-in-exile (the Gouvernement Royal Uni National du Kampuchéa, or GRUNK) had no armed forces, so the prince accepted a plan proposed by China and North Vietnam to ally with the Khmer Rouge, a quasi-communist rebel group, to gain a military foothold inside the country.137 The United States was the Khmer Rouge's primary opponent, providing military assistance to the Lon Nol government as part of a broader anti-communist strategy in Southeast Asia.138

Over the next several years, the Khmer Rouge consolidated its hold on the Cambodian countryside, instituting rigid social reforms, including collectivization.139 Relations between Sihanouk and the Khmer Rouge were strained, and Sihanouk's control over Khmer Rouge forces was often questioned.140

In 1973, Sihanouk demanded Cambodia's UN seat by both presenting credentials to the Credentials Committee and having allies request that the General Assembly vote on the matter. Sihanouk argued before the General Assembly that Khmer Rouge forces controlled 80 percent of Cambodia's territory and 90 percent of the population, but acknowledged that Lon Nol's forces held Phnom Penh and several other cities. The General Assembly elected not to render a decision on the issue, leaving Lon Nol's representatives in Cambodia's seat.141

In 1974, a second debate occurred in the General Assembly, resulting in a decision to revisit the issue in the next session.142 Although the Credentials Committee ultimately voted in December 1974 to accept the Lon Nol government's credentials, China (a Khmer Rouge ally at that point) voted in favor of rejection and three members abstained.143

Khmer Rouge forces continued to extend their control over Cambodian territory, and on January 1, 1975, began to bombard both Phnom Penh and the Pochentong airport.144 Throughout February and March, Khmer Rouge forces closed in on the capital. On March 30, the Soviet Union extended recognition to the Khmer Rouge/GRUNK coalition; and on April 1, President Lon Nol resigned and fled into exile, at which point India extended recognition.145 Khmer Rouge troops entered the outskirts of Phnom Penh in the first week of April, and the United States evacuated its embassy on April 12.146 France extended recognition on April 12, and on April 17, Khmer Rouge forces claimed full control over Phnom Penh.147 On April 18, China recognized the Khmer Rouge government; Canada on April 25; and Britain on May 2.148

The United States did not extend bilateral recognition to the Khmer Rouge regime, but also did not block its taking up Cambodia's UN seat.149 The Khmer Rouge representative received Cambodia's UN seat without comment at the beginning of the thirtieth session in October 1975.150

analysis. The Angolan case suggested that capital city control is necessary for large-scale bilateral recognition, even by allies. The Cambodian case affirms the conclusion that the capital is necessary, even when prior control has been established over the territory and general population. This requirement appears to hold for bilateral and multilateral recognition. The Cambodian case also demonstrates that the significance of capitals is not an African phenomenon. From the opponent perspective, the United States was again unique in withholding bilateral recognition from a government others had already recognized, suggesting a relatively strong commitment to the logic of self-interest, at least at the bilateral level. Other states, however, behaved as though the change in control over Phnom Penh was indeed decisive.

CHAD I, 1965–72 AND CHAD II, 1976–95

Where the Cambodian case illustrates the transfer of the capital at war's end, Chad provides several instances of recognition when combat was clearly ongoing. I note in the timeline section the three points at which recognition transfers occurred. Taken together, the first and second Chadian civil wars suggest that control of a capital alone can be sufficient for both levels of recognition, even by opponents, and that combatants can orient their campaigns to take this fact into account.

The Chad cases also provide some insight into the role of economic benefits for recognition. As a backdrop to the conflict, U.S. and French oil firms vied for the rights to explore Chad's potential oil reserves, and their national governments took this contest into account in their support of various factions.151

timeline, chad i. The first Chadian civil war began in 1965 with a rebellion against the government of Francis Tombalbaye, who had ruled Chad since its independence in 1960.152 In June 1966, a group calling itself the Front de Libération National (FROLINAT) began a guerrilla campaign in the countryside.153 The group received support initially from Sudan, but by 1969 Muammar Qaddafi had taken control of Libya and extended support to FROLINAT as well.154 FROLINAT made inroads against the Chadian government forces until Tombalbaye requested intervention from French troops in 1968, who slowed and reversed FROLINAT's advance.155

As part of a diplomatic spat with the Tombalbaye government, Qaddafi formally recognized FROLINAT as Libya's government and invited the group to train on Libyan territory in 1971.156 At the time Qaddafi extended recognition, FROLINAT's forces were confined to a single region (Tibesti) along the Libyan border.157 No other country or organization extended recognition to FROLINAT. By 1972, Tombalbaye had improved relations with Libya, and Qaddafi withdrew his recognition of FROLINAT.158

timeline, chad ii. In April 1975, Tombalbaye was assassinated by a group of junior military officers in a coup, and Félix Malloum became head of state.159 By 1976, the remnants of FROLINAT had launched campaigns against the new government. All of the FROLINAT factions, other than the one led by Hissène Habré, received support from Libya.160

Malloum and Habré signed a peace agreement in August 1978, but the agreement was short-lived. In February 1979, fighting between Habré's forces and Malloum's forces began in the capital, N'Djamena.161 Another former FROLINAT leader, Goukouni Oueddei, used the opportunity to push his forces into N'Djamena as well. By February 25, 1979, the conflict had caused more than 1,000 casualties, and 80,000 civilians had fled the city.162

When peace talks began, both Goukouni's and Habré's forces were present in the capital, but some observers suggest that Goukouni's forces had the advantage.163 Goukouni certainly benefited most from the setup of the coalition government that emerged when Malloum resigned. Although eleven factions were represented in the successor Gouvernement d'Union Nationale et de Transition (GUNT), Goukouni was named president and Habré became minister of defense.164

The GUNT was established as the government of Chad on November 11, 1979.165 Its level of control outside the capital was minimal, however.166 On November 14, the United Nations Credentials Committee accepted the credentials of Chad's representative without comment (Transfer 1).167

Goukouni and Habré quickly resumed their battle for control of N'Djamena.168 Habré's forces established control over one administrative zone in the city by the end of March 1980.169 Libyan forces intervened at Goukouni's request and succeeded in pushing Habré's forces out of their positions by December 1980.170 Habré fled to the east of the country.171 Estimates are that these two battles for N'Djamena led to 100,000 (of 200,000 total inhabitants at the time) displaced into Cameroon and destruction of most of the city's physical infrastructure.172

Habré regrouped with approximately $10 million in covert funds from the United States. France supported the GUNT openly with weapons shipments and endorsed the deployment of Organization of African Unity (OAU) forces to back-stop Goukouni's government.173 When OAU peacekeepers replaced the Libyans, however, Habré went on the offensive.174 His forces mounted a rapid assault on N'Djamena and took the city on June 7, 1982; GUNT forces withdrew north.175 Nigeria and Guinea extended recognition to Habré within days.176 The United States recognized Habré's regime later in the month.177 Habré's representative was seated at the UN without incident in October at the thirty-seventh session (Transfer 2).178 Libya (an opponent) delayed recognition until 1988.179

Habré's forces never had effective control over Chad, forcing the new leader to rely on continued support from both the United States and France, including the redeployment of French troops in 1983 to defend against Libyan-supported factions.180 On December 2, 1990, one such faction, led by Idriss Déby, succeeded in taking N'Djamena a day after Habré fled the city. The campaign was quick, as Déby's forces traveled 600 miles across the Sahel in two days.181 Déby officially proclaimed a new government on December 4.182 The French government began dealing with Déby as head of state within days, extending 90 million francs in credit and an invitation to Paris.183 The United States, which supported Habré's rule and initially considered Déby “a pro-Libyan puppet,” delayed recognition for four months.184 The Déby government's representative was accepted at the UN in November 1991 for the forty-sixth session without incident or discussion (Transfer 3).185 Déby has been the recognized head of government in Chad since 1990, though the second civil war continued until 1995.

analysis. The Chadian case illustrates the interplay of economic interests, recognition, and the military conflict. There was little question that any of the combatants controlled significant territory outside the capital. Sprints to N'Djamena were the preferred tactic for acquiring political authority.186 Multilateral recognition, once the capital was held, was uncontroversial at the UN, and the facts of (the lack of) control were unambiguous. Control of the capital appeared to be necessary, but also sufficient, not once, but on several occasions.

The Chadian case also illustrates the relative importance of economic interests. As competitors for the right to exploit Chad's potential oil wealth located primarily in the south of the country, the United States and France often supported rival groups. Each was quick to provide recognition to their preferred party upon capture of the capital, but not before. The only country deviating from this pattern was Libya, which recognized FROLINAT without its holding the capital and delayed recognition of the Habré government even when he did. Libya's lead was never followed, however. Instead, allies and the international community used capital city control as a necessary and sufficient condition; the United States, when it played the role of opponent, delayed bilateral recognition but not indefinitely.

The Chadian case also highlights the consequences of fixating on the capital: casualties and damage were concentrated in that space on multiple occasions, specifically to achieve control of the city with the ultimate end of gaining or maintaining international recognition.

Remaining Cases

Across the case studies, both self-interest and the capital control test explain some third-party recognition behavior. The United States, in particular, appears to grant or withhold bilateral recognition on the basis of self-interest. Other states, however, appear more willing to treat control of the capital as a necessary and sufficient condition for recognition. This section investigates how well the self-interest and capital city control models travel to the rest of the data (fifty-six wars with twelve recognition events).187

First, I consider how often allies prematurely recognize new rebel governments. Three of the four civil wars discussed in the case studies included premature recognition (Cambodia I was the exception). In the remaining data, premature recognition is also the norm; eleven out of the twelve recognition events qualify as premature. Of these premature recognitions, rebel groups held the capital in nine out of eleven cases.188 The median number of days between capital seizure and ally recognition was two.189

What about the “most likely” cases for a self-interested model of recognition, when economic gains are available? Do allies recognize prior to capital seizure (as with Libya), or do they wait (as with Chad II)? Of the fifty-six civil wars in the remaining data, twenty-four occurred in states with natural resources or petroleum reserves. Of these twenty-four, rebels gained control over those resources and had active allies in fourteen cases. Allies extended premature recognition to those rebels in only two cases and both times after the capital had changed hands. In other words, economic gain does not appear to be sufficient for recognition in the remainder of the data. The general pattern conforms to Chad II, not Libya.

How often did multilateral recognition follow who held the capital? In three of eleven cases, it did not. These three bilaterally recognized governments were either blocked from taking seats at the UN (Taliban/Afghanistan; People's Republic of Kampuchea/Cambodia II) or substantially delayed (People's Republic of China/China).190 While the Taliban and People's Republic of Kampuchea cases had broad international consensus for nonrecognition, the primary architect of the People's Republic of China/China delay was the United States.191 These exceptions are notable, but not the rule. In eight of the eleven cases, rebel groups that received bilateral recognition from allies and held their capital also received UN seats.

Given that the United States appears willing to delay multilateral recognition and to deny bilateral recognition, how often do other states do the same? Rarely. In the remaining cases, only in Tajikistan did an opponent (Afghanistan) deny bilateral recognition.192

On balance, then, whereas the self-interest model appears to fit the United States well, particularly with respect to its behavior as an opponent, the capital control test appears more consistent with behavior of states in general, both as allies and opponents.

If combatants expect that winning the capital will also win them the political prize of recognition, how often do cases resemble Chad II, where the capital is prioritized over establishing broader territorial control? The data from the remaining conflicts suggest that although rebel groups do not always have the option of prioritizing the capital, given territorial and personnel constraints, those that can and do subsequently improve their odds of ultimate victory. In sixteen of the fifty-six remaining conflicts, rebels were able to contest the capital at some stage and achieved widespread recognition seven times. Those rebel victories, however, came overwhelmingly from efforts that prioritized the capital (five out of seven wins), rather than through the gradual accumulation of territory.193 Rebels rarely win if they never contest the capital (four rebel victories in forty civil wars).

Finally, if prioritizing control of the capital appears to improve the odds of achieving political victory, what are the consequences? Of the sixteen conflicts in the remaining data in which the capital was contested, eleven resulted in significant destruction and/or displacement of civilians.194 Anecdotally, in Monrovia, Liberia, “all major infrastructures were damaged or looted,” as battles for the city occurred several times over the course of both Liberian civil wars and government employees fled.195 The Battle of Kabul (1992–96) was similarly catastrophic for residents and infrastructure.196 Accounts of the battle for Brazzaville in 1997 suggest that fighting over control of a capital city can be utterly destructive, even in a short time, with the right weaponry.197

Combatants also appear aware that the odds of winning politically rest on who holds the capital city. The fixation of rebel leader Charles Taylor on capturing Monrovia, for example, was not unique. The combatants in the early stages of Somalia's civil war exhibited the same awareness and tactical focus on controlling Mogadishu, which was equated early on with international recognition.198

Nor does seizing the capital foreshadow the end of fighting. In the twelve conflicts remaining in the data where capitals are seized and rebels receive bilateral recognition from at least one ally, their civil wars continue for approximately five years.199

It is important to note that combatants may quite reasonably equate their capitals with unique symbolic value in signaling leadership credibility to their domestic audiences.200 But capitals are not treated, even by domestic actors, as sufficient for claiming total victory—if that were the case, surrender of an incumbent would routinely follow capital seizure. Only the international community treats control of the capital as sufficient for political victory.

Conclusion

Civil wars weaken many state structures. The criteria states use to accord recognition during these wars has the potential to contribute to state weakness in several distinct ways. The first is by setting the bar low. Using a minimal standard of territorial control for recognition, the international community often creates governments whose authority is immediately in question. This low bar may also act as an incentive to initiate conflict rather than to pursue alternative forms of power transition such as electoral victory, though I did not test this hypothesis directly. State weakness is also exacerbated by the particular standard currently in use: privileging control of the capital puts a target on a developed region of the state and often on the highest concentration of its citizens. These assets can be destroyed and these people displaced as part of combatants' rational responses to strong incentives generated by the international community.

The criteria used by the international community to accord recognition during civil war have not been static, however. International legal scholarship has documented the changes to recognition criteria, both stated and enforced, over time.201 Territorial control has not always been a sufficient condition for recognition. States have also considered measures of a new government's legitimacy, including the willingness to conform to international financial and treaty obligations, the endorsement of the new executive by the legislature, and in some cases the results of popular referendums.202

In the past, states have also attempted to shift collective recognition standards to reduce the incentive for undesirable behavior. Both Central American leaders in the early twentieth century and African leaders in the early twenty-first have attempted to reject coups as a political transition mechanism by refusing to recognize in regional forums those who come to power that way.203 This is clearly a pragmatic practice for the leaders involved, but it also illustrates the kind of explicit coordination that can occur around recognition standards.

It seems unreasonable to expect civil war to vanish as a means of power transition. The primary concern for recognizers, therefore, should be altering the incentives that now encourage maximally destructive and costly tactics to achieve political victory.

Although no systematic change will affect the behavior of states that wish to accord recognition based on a self-interested calculus, the standards of multilateral recognition could be reconsidered. I suggest two changes to multilateral recognition that jointly could alter combatant incentives.

First, territorial control thresholds of any kind should be abandoned as a sufficient criterion for recognition. The Cambodia and Chad case studies illustrate why territorial control is not a Goldilocks problem. Any “just right” amount of territory—any level of the bar—creates the same incentive to inflict whatever damage is required to meet that threshold, if it is sufficient for recognition.

Second, the UN should begin to accept empty seats as a genuine alternative when representation disputes arise. Although this measure has been proposed within the Credentials Committee in several cases, it is almost never agreed to.204 A credible threat of nonrecognition—including of the incumbent government—and the corresponding loss of rights within institutions such as the World Bank would increase the collective leverage of member states over all combatants. Political victory, in the form of multilateral recognition, would cease to be zero sum. All combatants could lose.

These two changes to multilateral recognition open up the possibility of according recognition based on other criteria. Although the capital control test is relatively easy for third-party states to monitor, it is costly for civilians and, more abstractly, for the international system. Any new standard must weigh the cost of monitoring against potential gains. One possible criterion with relatively low incremental monitoring costs is a disqualification standard based on battlefield conduct. Data on combatant conduct are already collected by nongovernmental organizations and international bodies. Similar to coups disqualifying their perpetrators, certain actions in the course of war—indiscriminate targeting of civilians, ethnic cleansing, and the use of certain types of weaponry, for example—could disqualify combatants (the leadership group, in particular) as candidates for recognition. This would not guarantee good behavior by combatants, but it would increase the costs of bad behavior, possibly to prohibitive levels.

By setting the standards for what it means to be recognized as a government, third-party actors wield considerable influence over civil conflict without ever providing a single bullet or dollar. Altering these standards has the potential to affect the behavior of combatants during civil war. Ultimately, international recognition will be what states make of it.

Acknowledgments

For helpful comments during the development of this project, the author thanks Fotini Christia, Jeffry Frieden, Cullen Nutt, Roger Petersen, Barry Posen, Lucas Stanczyk, Lily Tsai, and panel participants at the 2017 annual meeting of the International Studies Association. The author also thanks the anonymous reviewers for their insightful suggestions.

Notes

1. 

Anthony Bell and David Witter, “The Libyan Revolution: Roots of Rebellion” (Washington, D.C.: Institute for the Study of War, 2011), p. 8, http://www.understandingwar.org/report/libyan-revolution-part-i-roots-rebellion.

2. 

Stefan A.G. Talmon, “Recognition of the Libyan National Transitional Council,” ASIL Insight Vol. 1, No. 6 (2011), p. 3.

3. 

For France and the United Arab Emirates, see ibid., pp. 3–4. For the United States, see U.S. Department of State, “Special Briefing, 15 July 2011” (Washington, D.C.: U.S. Department of State, 2011), https://2009-2017.state.gov/r/pa/prs/ps/2011/07/168662.htm. For the United Kingdom, see Colin Warbrick, “British Policy and the National Transitional Council of Libya,” International and Comparative Law Quarterly, Vol. 61, No. 1 (January 2012), pp. 252–253, doi:10.1017/S0020589311000649. For the Libya Contact Group generally, see United Nations (UN) Security Council, “Libya: August 2011 Monthly Forecast” (New York: UN, 2011), http://www.securitycouncilreport.org/monthly-forecast/2011-08/.

4. 

For assets held in Italy, see Osvaldo Croci and Marco Valigi, “Continuity and Change in Italian Foreign Policy: The Case of the International Intervention in Libya,” Contemporary Italian Politics, Vol. 5, No. 1 (April 2013), pp. 38–54, doi:10.1080/23248823.2013.781373. For assets held in the United Kingdom, see Chriso Summers, “How Did Libyan Money Come to Be Printed in Britain?” BBC News, September 2, 2011, http://www.bbc.co.uk/news/uk-14746873. For the U.S. case, see Sebnem Arsu and Steven Erlanger, “Libya Rebels Get Formal Backing, and $30 Billion,” New York Times, July 15, 2011.

5. 

Louis Charbonneau and Hamuda Hassan, “France Defends Arms Airlift to Libyan Rebels,” Reuters, June 29, 2011, https://www.reuters.com/article/us-libya/france-defends-arms-airlift-to-libyan-rebels-idUSTRE7270JP20110629.

6. 

Sam Foster Halabi, “Traditions of Belligerent Recognition: The Libyan Intervention in Historical and Theoretical Context,” American University International Law Review, Vol. 27, No. 2 (2012), pp. 381–382; and Julian Borger and Terry Macalister, “The Race Is on for Libya's Oil, with Britain and France Both Staking a Claim,” Guardian, September 1, 2011, http://www.theguardian.com/world/2011/sep/01/libya-oil.

7. 

Anthony Bell, Spencer Butts, and David Witter, “The Libyan Revolution: The Tide Turns” (Washington, D.C.: Institute for the Study of War, 2011), pp. 17–20, http://www.understandingwar.org/report/libyan-revolution-part-4-tide-turns.

8. 

A list of individual recognition statements of the NTC is available in the online supplementary material, doi:10.7910/DVN/72GCGD. I have followed Talmon, “Recognition of the Libyan National Transitional Council,” in coding as recognition of the NTC only those statements in which the NTC is recognized as the government or sovereign authority, not as simply the representative of the Libyan people.

9. 

Steve Gutterman, “Russia Recognizes Libya's Transitional Council,” Reuters, September 1, 2011, https://www.reuters.com/article/us-libya-russia/russia-recognises-libyas-transitional-council-idUSTRE7800WS20110901; and Tania Branigan, “Libya's NTC Gains Chinese Recognition” Guardian, September 12, 2011, https://www.theguardian.com/world/2011/sep/12/libya-ntc-chinese-recognition.

10. 

Twelve states protested this action in the UN General Assembly. See UN, “After Much Wrangling, General Assembly Seats National Transitional Council of Libya as Country's Representative for Sixty-Sixth Session (GA/11137)” (New York: UN, September 16, 2011), http://www.un.org/News/Press/docs/2011/ga11137.doc.htm.

11. 

The legal significance of recognition has been a subject of study in Western international law since Hugo Grotius (1583–1645). Canonical examples include Hersch Lauterpacht, Recognition in International Law, Vol. 3 (1947; repr. Cambridge: Cambridge University Press, 2013); Chen Ti-Chiang, The International Law of Recognition, with Special Reference to Practice in Great Britain and the United States, L.C. Green, ed. (New York: Praeger, 1951); Hans Kelsen, “Recognition in International Law: Theoretical Observations,” American Journal of International Law, Vol. 35, No. 4 (October 1941), pp. 605–617, doi:10.2307/2192561; and Quincy Wright, “Some Thoughts about Recognition,” American Journal of International Law, Vol. 44, No. 3 (July 1950), pp. 548–559, doi:10.2307/2194030. For Cold War and post–Cold War scholarship, see L. Thomas Galloway, Recognizing Foreign Governments: The Practice of the United States (Washington, D.C.: American Enterprise Institute, 1978); M.J. Peterson, Recognition of Governments: Legal Doctrine and State Practice, 1815–1995 (New York: St. Martin's, 1997); and Brad R. Roth, Governmental Illegitimacy in International Law (Oxford: Oxford University Press, 2000).

12. 

Christopher S. Chivvis and Jeffrey Martini, Libya after Qaddafi: Lessons and Implications for the Future (Washington, D.C.: RAND Corporation, 2014), pp. 20–24.

13. 

Bell and Witter, “The Libyan Revolution: Roots of Rebellion,” pp. 20–21. From 2011 through 2014, the average number of battle deaths per year was far more than the 100 battle deaths per year that qualify a war as ongoing under the coding rules used in this analysis. See “UCDP Battle-Related Deaths Dataset v.5-2015” (Uppsala, Sweden: Department of Peace and Conflict Research, Uppsala University, 2015).

14. 

For example, the UN Support Mission in Libya was established on September 16, 2011, although its footprint was limited. See Chivvis and Martini, Libya after Qaddafi, p. 6. In December 2011, the NTC also received a peacebuilding grant of $3 million from the World Bank, which lends only to the sovereign governments of member countries. All UN members are members of the International Monetary Fund and the World Bank Group, with the exception of Andorra, Cuba, Liechtenstein, Monaco, and North Korea. See World Bank Treasury, “Everything You Always Wanted to Know about the World Bank” (Washington, D.C.: World Bank, n.d.), http://treasury.worldbank.org/cmd/pdf/WorldBankFacts.pdf; and Simon Gray to Hassan Ziglam, December 20, 2011, World Bank Documents and Reports online archive, http://documents.worldbank.org/curated/en/614991468046737022/Official-Documents-Agreement-for-Grant-TF011413.

15. 

George Modelski, “The International Relations of Internal War,” in James N. Rosenau, ed., International Aspects of Civil Strife (Princeton, N.J.: Princeton University Press, 1964), p. 31. Multilateral recognition usually occurs when a meeting of member-state representatives convenes, and so the timing is less discretionary than the timing of bilateral recognition. Some bodies, such as the UN, convene regularly on established timetables (see below for the discussion of UN credentialing). Other bodies convene irregularly.

16. 

If the analysis is restricted to those cases where recognition occurred at least one year before war's end, then 14 of 18 cases would be considered premature. Of the 2 cases in which recognition occurred within one year of the war's end, 1 (Ethiopia) has a 365-day spread between first bilateral recognition and the end of the war. The other (China) has a 211-day spread, and the analysis assumes that Taiwan is not within the state's borders.

17. 

The mean is 9.7 years. The distribution of conflict duration after premature recognition has a long tail, so I report the median unless noted otherwise.

18. 

Lauterpacht, Recognition in International Law, Vol. 3, p. 125; and Roth, Governmental Illegitimacy in International Law, p. 136.

19. 

I follow Stephen D. Krasner's definition of “rule” here. See Krasner, “Structural Causes and Regime Consequences: Regimes as Intervening Variables,” International Organization, Vol. 36, No. 2 (Spring 1982), p. 186, doi:10.1017/S0020818300018920. See also Michael Barnett and Martha Finnemore, Rules for the World: International Organizations in Global Politics (Ithaca, N.Y.: Cornell University Press, 2004).

20. 

For Mogadishu, see Isabelle Duyvesten, Clausewitz and African War: Politics and Strategy in Liberia and Somalia (London: Frank Cass, 2005), p. 113; John Drysdale, Whatever Happened to Somalia? 2nd ed. (London: HAAN, 2001), p. x; and Anna Simons, Networks of Dissolution: Somalia Undone (Boulder, Colo.: Westview, 1995), p. 91. For Brazzaville, see UN Security Council, “Report of the Secretary-General on the Situation in the Republic of Congo (S/1997/814)” (New York: UN, October 21, 1997), pp. 7–8. For Kabul, see Human Rights Watch, Blood-Stained Hands: Past Atrocities in Kabul and Afghanistan's Legacy of Impunity (New York: Human Rights Watch, 2005), pp. 22–100, http://www.hrw.org/reports/2005/afghanistan0605/afghanistan0605.pdf; and Richard S. Newell, “Chapter 4: Government and Politics,” in Peter Blood, ed., Afghanistan: A Country Study (Washington, D.C.: Library of Congress, n.d.), http://www.country-data.com/frd/cs/afghanistan/afghanistan.html.

21. 

William Reno, “How Sovereignty Matters: International Markets and the Political Economy of Local Markets in Weak States,” in Thomas Callaghy, Ronald Kassimir, and Robert Latham, eds., Intervention and Transnationalism in Africa: Global-Local Networks of Power (Cambridge: Cambridge University Press, 2001), p. 202.

22. 

Paul Collier et al., Breaking the Conflict Trap: Civil War and Development Policy (Washington, D.C.: World Bank, 2003), p. 15.

23. 

State strength has many definitions. I rely on Robert I. Rotberg's premise that state strength should be understood as a function of the ability to effectively deliver vital public goods. See Rotberg, “Failed States, Collapsed States, Weak States: Causes and Indicators,” in Rotberg, ed., State Failure and State Weakness in a Time of Terror (Washington, D.C.: Brookings Institution Press, 2003), p. 3. Similarly, the U.K. Department for International Development defines state fragility based on an inability to deliver core functions to a majority of the population. See Department for International Development, “Why We Need to Work More Effectively in Fragile States” (London: Department for International Development, 2005), p. 7.

24. 

Peterson, Recognition of Governments, p. 29.

25. 

Roth, Governmental Illegitimacy in International Law, p. 135.

26. 

Restatement of the Law, Third: Foreign Relations Law of the United States, Vol. 1 (St. Paul: American Law Institute, 1987), section 203.a.

27. 

Peterson, Recognition of Governments, p. 2. For a discussion of the distinction (or lack thereof) between the phrases “de facto recognition” and “de jure recognition,” see Lauterpacht, Recognition in International Law, Vol. 3, p. 284–288. Peterson clarifies that de jure recognition is the default position unless otherwise specified. See Peterson, Recognition of Governments, p. 86. U.S. recognition of the first government of Israel is an example of this kind of specification. President Harry Truman recognized “the provisional Government as the de facto authority of the new state of Israel” on May 14, 1948. See Statement of the President re: Recognition of Israel, May 14, 1948, Handwriting of the President, Alphabetical Correspondence File, 1916–1950, Charles G. Ross Papers, Harry S. Truman Library and Archives, Independence, Missouri. The recognition was intended as a placeholder until a permanent government was elected. After elections in January 1949, Truman extended de jure recognition to the resulting entity. See Press Release, January 31, 1949, OF 204-D: Jewish State, 1948–1948, box 914, Official File, Truman Papers, Harry S. Truman Library and Archives. This type of nuanced recognition is rare after 1945.

28. 

Restatement of the Law, Third, section 203.b.

29. 

Wright, “Some Thoughts about Recognition,” p. 555.

30. 

M.J. Peterson, “Political Use of Recognition: The Influence of the International System,” World Politics, Vol. 34, No. 3 (April 1982), p. 325, doi:10.2307/2010322.

31. 

Ibid., p. 329. Recognition of the NTC in Libya qualified as premature. See Talmon, “Recognition of the Libyan National Transitional Council,” p. 5; and John B. Bellinger III, “Legal Questions in U.S. Nod to Libya's Opposition” (New York: Council on Foreign Relations, July 18, 2011).

32. 

Lauterpacht, Recognition in International Law, Vol. 3, pp. 93–94.

33. 

Charles Zorgbibe, La Guerre Civile [Civil war] (Paris: Presses Universitaires de France, 1975), p. 142.

34. 

For a specific application of this argument, see Pierre Englebert, “Why Congo Persists: Sovereignty, Globalization, and the Violent Reproduction of a Weak State” (Oxford: Queen Elizabeth House, Oxford University, February 2003). For the alternative argument that recognition is primarily a declaratory act (i.e., an observation of fact), see Thomas D. Grant, The Recognition of States: Law and Practice in Debate and Evolution (Westport, Conn.: Praeger, 1999).

35. 

Peterson, Recognition of Governments, pp. 144–152. Some states try to avoid the recognition issue by arguing that their conduct is sufficient to signal recognition of the entity in question, without a formal announcement. Stefan Talmon documents this for the United Kingdom since 1980. See Talmon, “Recognition of Governments: An Analysis of the New British Policy and Practice,” British Yearbook of International Law, Vol. 63, No. 1 (January 1993), pp. 231–297, doi:10.1093/bybil/63.1.231. Peterson describes this as “tacit recognition” and notes that certain behaviors (establishment of formal diplomatic relations or formal consular relations, or conclusion of a bilateral treaty) have long been considered equivalent to formal explicit recognition. See Peterson, Recognition of Governments, pp. 86–88. I consider them equivalent here.

36. 

Hans Aufricht, “Principles and Practices of Recognition by International Organizations,” American Journal of International Law, Vol. 43, No. 4 (October 1949), pp. 699–704, doi:10.2307/2193259.

37. 

Farrokh Jhabvala, “The Credentials Approach to Representation Questions in the UN General Assembly,” California Western International Law Journal, Vol. 7 (Summer 1977), pp. 615–636; and Suellen Ratliff, “UN Representation Disputes: A Case Study of Cambodia and a New Accreditation Proposal for the Twenty-first Century,” California Law Review, Vol. 87, No. 5 (October 1999), pp. 1207–1264, doi:10.15779/Z38C71W. There was some debate on this point at the UN, particularly in 1950 as related to the representation of China, but legal scholars support this interpretation. See Hans Kelsen, The Law of the United Nations: A Critical Analysis of the Fundamental Problemswith Supplement (New York: Praeger, 1950), pp. 946–947.

38. 

Kelsen, The Law of the United Nations, p. 947 n. 4.

39. 

Peterson, Recognition of Governments, pp. 103–120.

40. 

Talmon, “Recognition of the Libyan National Transitional Council,” pp. 2–3; and Stefan Talmon, “Recognition of Opposition Groups as the Legitimate Representative of a People,” Chinese Journal of International Law, Vol. 12 (2013), pp. 226–230, doi:10.1093/chinesejil/jmt014.

41. 

Christopher J. Le Mon, “Unilateral Intervention by Invitation in Civil Wars: The Effective Control Test Tested,” NYU Journal of International Law & Politics, Vol. 35 (2002), p. 743.

42. 

For Angola and oil revenues, see George Wright, The Destruction of a Nation: United States' Policy toward Angola since 1945 (Chicago: Pluto, 1997), p. 72. For a discussion of ship registration revenues in the Liberia case, see Michael E. Field, “Liberia v. Bickford: The Continuing Problem of Recognition of Governments and Civil Litigation in the United States,” Maryland Journal of International Law & Trade, Vol. 18, No. 1 (1994), pp. 113–148.

43. 

Bellinger, “Legal Questions in U.S. Nod to Libya's Opposition”; and Talmon, “Recognition of Opposition Groups as the Legitimate Representative of a People,” p. 244.

44. 

For an example in Chad, see William J. Foltz, “Chad's Third Republic: Strengths, Problems, and Prospects,” CSIS Africa Notes, October 30, 1987, p. 2. For an example in Liberia, see UN Security Council, Resolution 1509 (New York: UN, September 19, 2003).

45. 

See James R. Crawford, The Creation of States in International Law (Cambridge: Cambridge University Press, 2006); Bridget 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, doi:10.1017/S0020818311000105; Mikulas Fabry, Recognizing States: International Society and the Establishment of New States since 1776 (Oxford: Oxford University Press, 2010); and Bridget Coggins, Power Politics and State Formation in the Twentieth Century: The Dynamics of Recognition (Cambridge: Cambridge University Press, 2014).

46. 

James Crawford defines secession as “the creation of a State by the use or threat of force without the consent of the former sovereign.” See Crawford, The Creation of States in International Law, p. 375.

47. 

Ibid., p. 376.

48. 

Fabry, Recognizing States, pp. 147–149.

49. 

Crawford, The Creation of States in International Law, p. 390.

50. 

Coggins, “Friends in High Places,” pp. 435–436.

51. 

Ibid., p. 436.

52. 

Bridget L. Coggins, “Rebel Diplomacy: Theorizing Violent Non-State Actors' Strategic Use of Talk,” in Ana Arjona, Nelson Kasfir, and Zachariah Mempilly, eds., Rebel Governance in Civil War (Cambridge: Cambridge University Press, 2015), pp. 98–118; and Reyko Huang, “Rebel Diplomacy in Civil War,” International Security, Vol. 40, No. 4 (Spring 2016), pp. 89–126, doi:10.1162/ISEC_a_00237.

53. 

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

54. 

Reno, “How Sovereignty Matters,” pp. 207–208.

55. 

For more on the “booty futures” concept, see Michael L. Ross, “How Do Natural Resources Influence Civil War? Evidence from Thirteen Cases,” International Organization, Vol. 58, No. 1 (Winter 2004), pp. 58–59, doi:10.1017/S002081830458102X. For the Chadian case, see Macartan Humphreys, “Natural Resources, Conflict, and Conflict Resolution: Uncovering the Mechanisms,” Journal of Conflict Resolution, Vol. 49, No. 4 (August 2005), pp. 508–509, doi:10.1177/0022002705277545; and Millard J. Burr and Robert O. Collins, Africa's Thirty Years' War: Libya, Chad, and the Sudan, 1963–1993 (Boulder, Colo.: Westview, 1999).

56. 

Reno, “How Sovereignty Matters,” p. 199; Felix Gerdes, Civil War and State Formation: The Political Economy of War and Peace in Liberia (Frankfurt-am-Main: Campus Verlag, 2013), p. 111.

57. 

Odd Arne Westad, The Global Cold War: Third World Interventions and the Making of Our Times (Cambridge: Cambridge University Press, 2005), p. 236.

58. 

Huang, “Rebel Diplomacy in Civil War,” pp. 97–100.

59. 

Modelski, “The International Relations of Internal War,” p. 31.

60. 

Reno, “How Sovereignty Matters.”

61. 

Barack Obama, David Cameron, and Nicholas Sarkozy, “Opinion: Libya's Pathway to Peace,” New York Times, April 14, 2011, http://www.nytimes.com/2011/04/15/opinion/15iht-edlibya15.html.

62. 

For example, the NTC was displeased about Russia's relatively slow recognition of it. See Federica Saini Fasanotti, “Russia and Libya: A Brief History of an On-Again-Off-Again Friendship,” Order from Chaos blog, September 1, 2016, https://www.brookings.edu/blog/order-from-chaos/2016/09/01/russia-and-libya-a-brief-history-of-an-on-again-off-again-friendship.

63. 

For examples, see Amikam Nachmani, International Intervention in the Greek Civil War: The United Nations Special Committee on the Balkans, 1947–1952 (New York: Greenwood, 1990), p. 113; and Odd Arne Westad, “Introduction,” in Westad, ed., Brothers in Arms: The Rise and Fall of the Sino-Soviet Alliance, 1945–1963 (Palo Alto, Calif.: Stanford University Press, 1998), p. 8.

64. 

Coggins, “Friends in High Places,” pp. 451–453.

65. 

See Marika Landau-Wells, “Capital Cities in Civil Wars: The Locational Dimension of Sovereign Authority” (London: Crisis States Research Centre, 2008). This position is also already embedded in some political science scholarship by virtue of dataset coding. For example, the Uppsala Conflict Data Program/Peace Research Institute Oslo (UCDP/PRIO) Armed Conflict Dataset defines the government side of a civil conflict as the entity holding the capital. See “UCDP/PRIO Armed Conflict Dataset Codebook, Version 4-2015” (Uppsala, Sweden: Department of Peace and Conflict Research, Uppsala University, 2015), p. 2, http://www.pcr.uu.se/research/ucdp/datasets/ucdp_prio_armed_conflict_dataset; 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, doi:10.1177/0022343302039005007.

66. 

Peterson, Recognition of Governments, pp. 173–174.

67. 

Jeffrey Herbst, States and Power in Africa: Comparative Lessons in Authority and Control (Princeton, N.J.: Princeton University Press, 2000), p. 111. For a discussion of the mismatch between the empirical state and its territorial jurisdiction, see Stephen D. Krasner, Sovereignty: Organized Hypocrisy (Princeton, N.J.: Princeton University Press, 1999).

68. 

Robert H. Jackson, “Juridical Statehood in Sub-Saharan Africa,” Journal of International Affairs, Vol. 46, No. 1 (Summer 1992), p. 8. For a discussion of remote resources, see Philippe Le Billon, “Angola's Political Economy of War: The Role of Oil and Diamonds, 1975–2000,” African Affairs, January 2001, pp. 55–80, doi:10.1093/afraf/100.398.55; and Humphreys, “Natural Resources, Conflict, and Conflict Resolution.” For the significance of capitals generally, see Martin Glassner and Chuck Fahrer, Political Geography, 3rd ed. (New York: John Wiley, 2004); and John Taylor, Jean G. Lengellé, and Caroline Andrew, eds., Capital Cities: International Perspectives (Ottawa: Carleton University Press, 1993).

69. 

M.J. Peterson, “Recognition of Governments Should Not Be Abolished,” American Journal of International Law, Vol. 77, No. 1 (January 1983), p. 33, doi:10.2307/2201197.

70. 

Daniel Kahneman and Shane Frederick, “Representativeness Revisited: Attribute Substitution in Intuitive Judgment,” in Thomas Gilovich, Dale W. Griffin, and Daniel Kahneman, eds., Heuristics and Biases: The Psychology of Intuitive Judgment (Cambridge: Cambridge University Press, 2002), pp. 49–81; and Robin M. Hogarth and Natalia Karelaia, “Heuristic and Linear Models of Judgment: Matching Rules and Environments,” Psychological Review, Vol. 114, No. 3 (July 2007), pp. 733–758, doi:10.1037/0033-295X.114.3.733. It is also consistent with everyday political metonymy (e.g., “Berlin says” as a proxy for “The government representing the people of Germany says”).

71. 

For a discussion of the issue, see Nicholas Sambanis, “What Is Civil War? Conceptual and Empirical Complexities of an Operational Definition,” Journal of Conflict Resolution, Vol. 48, No. 6 (December 2004), pp. 814–858, doi:10.1177/0022002704269355; and Christopher Blattman and Edward Miguel, “Civil War,” Journal of Economic Literature, Vol. 48, No. 1 (March 2010), pp. 3–57, doi:10.1257/jel.48.1.3.

72. 

For a discussion of the “sovereignty rupture” perspective in relation to the civil war literature, see Nicholas Sambanis and Jonah Schulhofer-Wohl, “Civil War as Sovereignty Rupture: Coding Intra-State Conflict, 1945–2015,” University of Pennsylvania and University of Virginia, June 22, 2015.

73. 

For recent dyadic analyses, see 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, doi:10.1177/0022002709336458; and Barbara F. Walter, “Why Bad Governance Leads to Repeat Civil War,” Journal of Conflict Resolution, Vol. 59, No. 7 (October 2015), pp. 1242–1272, doi:10.1177/0022002714528006.

74. 

Sambanis and Schulhofer-Wohl, “Civil Wars as Sovereignty Rupture,” p. 6.

75. 

James D. Fearon, “Why Do Some Civil Wars Last So Much Longer Than Others?” Journal of Peace Research, Vol. 41, No. 3 (May 2004), p. 278, doi:10.1177/0022343304043770. Cases enter this dataset when the state is a participant and at least 100 deaths occur on both sides each year on average, with at least 1,000 total battle deaths over the course of the conflict.

76. 

Coups are excluded because they are designed to present a fait accompli of effective control. For an explanation of this logic, see Modelski, “The International Relations of Internal War,” p. 27. Post-coup recognition hinges on whether states view coups as a legitimate means of political transition. Recognition of popular revolutions is similarly based on specific claims about the legitimacy of the method of transition. For a discussion, see Richard A. Falk, “Janus Tormented: The International Law of Internal War,” in Rosenau, International Aspects of Civil Strife, p. 199. The legal consequences of coups and popular revolutions can be very different from those of civil wars if the formal transition of political authority in the coup or revolution is brought about through resignation of the previous government rather than conquest or negotiated settlement. See A.C. Bundu, “Recognition of Revolutionary Authorities: Law and Practice of States,” International and Comparative Law Quarterly, Vol. 27, No. 1 (1978), pp. 18–45. In practice, these exclusions mean that most wars in the dataset start on the peripheries, far from the capital cities of the states in question. Of the “indeterminate” cases, I retained Burma (1948–) for the communist faction only; Colombia (1963–) for the original parties; Djibouti (1991–94); Ethiopia (1974–1992); Indonesia (1958–60); Nicaragua (1981–88); and Somalia (1991–2012). I exclude Chad (1994–98) as an attempted extortion by rebels in the South; Democratic Republic of Congo (1960–65), which was largely secessionist and where the power transfers occurred by coup; Pakistan (1993–99) as political violence and intimidation rather than civil war; Sudan (1963–72) because the primary aim was secession for the south; and Sudan (1983–) because the primary aim is restoring southern autonomy.

77. 

“UCDP Battle-Related Deaths Dataset v.5-2015”; Gleditsch et al., “Armed Conflict 1946–2001”; and Therése Pettersson and Peter Wallensteen, “Armed Conflicts, 1946–2014,” Journal of Peace Research, Vol. 52, No. 4 (July 2015), pp. 536–550, doi:10.1177/0022343315595927. Where datasets differ, I attempt to reconcile all conflicts back to Fearon's measures of onset and duration. War termination is coded for the year after which battle deaths decline below 100 for 24 consecutive months using UCDP data.

78. 

I recode Afghanistan I and II into a single conflict, because violence continued through 1996 following the installation of the mujahideen government in April 1992 as Gulbuddin Hekmatyar attempted to seize Kabul. See Human Rights Watch, Blood-Stained Hands, pp. 22–100. I recode the First and Second Congo Wars into a single conflict, because fewer than 24 months elapsed between Laurent Kabila's declaration of victory and the resumption of violence involving some parties from the original conflict. See Gérard Prunier, Africa's World War: Congo, the Rwandan Genocide, and the Making of a Continental Catastrophe (Oxford: Oxford University Press, 2008).

79. 

This case disappears from datasets that require one party to the conflict to be the recognized government.

80. 

Appendix A is available in the online supplementary materials.

81. 

A bibliography of each case is available in the online supplementary materials. For many cases, several sources were used. The author is particularly indebted to Mark J. Mullenbach, “Dynamic Analysis of Dispute Management (DADM) Project: Intrastate Dispute Narratives” (Conway: Department of Political Science, University of Central Arkansas, n.d.), http://uca.edu/politicalscience/dadm-project/dadm-intratstate-dispute-narratives/. For a discussion of process tracing, see Andrew Bennett, “Process Tracing and Causal Inference,” in Henry E. Brady and David Collier, eds., Rethinking Social Inquiry: Diverse Tools, Shared Standards, 2nd ed. (Lanham, Md.: Rowman and Littlefield, 2010), pp. 207–220.

82. 

Appendix B is available in the online supplementary materials. Recognitions of construct governments are excluded from the data. Two examples of construct governments are the Interim Government of National Unity in Liberia and the Transitional National Government in Somalia. These governing authorities were constituted explicitly by outsiders to fill a vacuum of authority, and the individuals making up these bodies were not rebel combatants.

83. 

George B. Davis, The Elements of International Law, with an Account of Its Origin, Sources, and Historical Development, a New and Revised Edition (New York: Harper and Bros., 1900), p. 276.

84. 

In theory, it is possible that allies providing covert assistance might go undetected. The scale of these conflicts, however (as determined by battle-death thresholds), ensures that they have been examined by a number of scholars and journalists. In most cases, third-party participation is well documented. Should a state's role in the conflict go undetected, however, the status would be incorrectly coded as neutral.

85. 

Wright, “Some Thoughts about Recognition,” pp. 552–553.

86. 

For a discussion of these acts, see Peterson, Recognition of Governments, pp. 86–88.

87. 

For detailed discussion of these distinctions in the Libyan and Syrian civil wars, see Talmon, “Recognition of Opposition Groups as the Legitimate Representative of a People,” pp. 219–253.

88. 

Often these clusters of recognitions are mentioned in secondary sources that do not disaggregate the recognizers; the quality of data from these sources is thus relatively low.

89. 

This focus on the UN omits the type of multilateral recognition provided by regional organizations such as the Association of Southeast Asian Nations or the African Union. This omission occurs, in part, because of space limits. Preliminary research by the author suggests that regional bodies are highly attuned to their constitutive power with respect to member governments and operate in a more strategic manner than does the UN when considering recognition. Given the international significance of recognition at the UN, however, that body is the focus of this analysis.

90. 

In 1950, following the first debate over which government (the Communist or Nationalist) was entitled to represent China, the General Assembly adopted Resolution 396 (V). The resolution stated that controversies over representation should be considered “in the light of the Purposes and Principles of the Charter.” For a discussion of the intentional vagueness of this wording, see Jhabvala, “The Credentials Approach to Representation Questions in the UN General Assembly,” pp. 629–633.

91. 

UN General Assembly, “Rules of Procedure of the General Assembly (Res/A/520/Rev.17)” (New York: UN, September, 2007), Rule 27.

92. 

Jhabvala, “The Credentials Approach to Representation Questions in the UN General Assembly,” pp. 619–620 n. 17.

93. 

Matthew Griffin, “Accrediting Democracies: Does the Credentials Committee of the United Nations Promote Democracy through Its Accreditation Process, and Should It?” New York University Journal of International Law & Politics, Vol. 32, No. 3 (Spring 2000), p. 731.

94. 

These protests are recorded in the reports of the Credentials Committee, traditionally agenda item 3 in a UN Assembly session.

95. 

Griffin, “Accrediting Democracies,” p. 733 n. 24.

96. 

Most objections regarding the acceptance of the reports of the Credentials Committee concern recognition of the State of Israel. I ignore such objections here.

97. 

All Credentials Committee reports were accessed using the UN Official Document System, https://documents.un.org/prod/ods.nsf/home.xsp. UN Yearbooks were used to identify the specific document symbol for each year's report. See http://unyearbook.un.org/.

98. 

These cases are not considered wars of secession, because there was international consensus that each new state should exist. The precise identity of the legitimate government at the time of the new state's accession into the UN was debated, however.

99. 

See the bibliography in the online supplementary materials for specific references.

100. 

One exception to this is the China case. The recognized Nationalist government lost its capital (Nanjing) on April 23, 1949, but the Communists sought international recognition only for a government based in Beijing (historically recognized as a capital of China) on October 1. Control of the capital is thus coded as starting on October 1, 1949. See David C. Wolf, “‘To Secure a Convenience’: Britain Recognizes China—1950,” Journal of Contemporary History, Vol. 18, No. 2 (April 1983), p. 313, doi:10.1177/002200948301800207.

101. 

This includes Charles Taylor's incursions into Monrovia from 1990 to 1992 and the partition of Beirut, Lebanon, between 1988 and 1990 (Lebanon II).

102. 

For oil and natural gas reserves, I use the coding in Humphreys, “Natural Resources, Conflict, and Conflict Resolution.” For non-petroleum natural resources, I use the coding in Halvard Buhaug and Scott Gates, “The Geography of Civil War,” Journal of Peace Research, Vol. 39, No. 4 (July 2002), pp. 417–433, doi:10.1177/0022343302039004003.

103. 

In several cases—Congo-Brazzaville, for example—it is clear that companies pursue their own agendas, independent of their national government and sometimes to that government's detriment. I do not code those cases as having benefits for the recognizing state. See Global Witness, “Time for Transparency: Coming Clean on Oil, Mining, and Gas Revenues” (Washington, D.C.: Global Witness, 2004), http://www.globalwitness.org/library/time-transparency.

104. 

Portugal was the target of sanctions for its resistance to decolonization and conduct of the Colonial War. For the British treatment of the Portuguese regime, see Norrie MacQueen and Pedro Aires Oliveira, “‘Grocer Meets Butcher’: Marcello Caetano's London Visit of 1973 and the Last Days of Portugal's Estado Novo,” Cold War History, Vol. 10, No. 1 (2010), pp. 29–50, doi:10.1080/14682740902764551.

105. 

Westad, The Global Cold War, p. 220.

106. 

“Angola—Developments Leading Up to Independence,” Keesing's Contemporary Archives, Vol. 21 (1975), pp. 27497–27498.

107. 

Westad, The Global Cold War, pp. 212–214, 218.

108. 

Memorandum of Conversation, June 20, 1975, Foreign Relations of the United States (FRUS) 1969–1976: Vol. 28: Southern Africa (Washington, D.C.: Government Printing Office [GPO], 2011), pp. 258–261; and Memorandum of Conversation, July 18, 1975, 9:01–10:12 a.m., FRUS 1969–1976, Vol. 28, p. 286.

109. 

Westad, The Global Cold War, pp. 227–235.

110. 

Minutes of a National Security Meeting, June 27, 1975, 2:30–3:20 p.m., FRUS 1969–1976, Vol. 28, pp. 266–270.

111. 

Ibid., p. 271.

112. 

Ibid., p. 266.

113. 

Paper Submitted to the 40 Committee, n.d., FRUS 1969–1976, Vol. 28, p. 282.

114. 

“Angola—Developments Leading Up to Independence,” p. 27498.

115. 

U.S. Department of State to All Posts, “Angolan Recognition,” November 8, 1975, document 1975STATE265503, Electronic Telegrams 1975, Central Foreign Policy Files created 7/1/1973–12/31/1979, Record Group 59, National Archives, Washington, D.C., http:/./aad.archives.gov/aad/.

116. 

U.S. Department of State, American Embassy Khartoum to Secretary of State, “Angolan Recognition,” November 10, 1975, document 1975KHARTO02618, Electronic Telegrams 1975, Central Foreign Policy Files created 7/1/1973-12/31/1979, Record Group 59, National Archives.

117. 

U.S. Department of State, American Embassy Paris to Secretary of State, “Angola Recognition,” November 10, 1975, document 1975PARIS29255, Electronic Telegrams 1975, Central Foreign Policy Files created 7/1/1973–12/31/1979, Record Group 59, National Archives.

118. 

U.S. Department of State, American Embassy Dar es Salaam to Secretary of State, “Angolan Recognition,” November 9, 1975, document 1975DARES05045, Electronic Telegrams 1975, Central Foreign Policy Files created 7/1/1973–12/31/1979, Record Group 59, National Archives.

119. 

U.S. Deptartment of State, American Embassy Ottawa, to Secretary of State, “Angolan Recognition,” November 10, 1975, document 1975OTTAWA04215, Electronic Telegrams 1975, Central Foreign Policy Files created 7/1/1973–12/31/1979, Record Group 59, National Archives.

120. 

“Angola—Developments Leading Up to Independence,” p. 27497.

121. 

Telegram from the Department of State to All Diplomatic Posts, November 13, 1975, 11:07 p.m., FRUS 1969–1976, Vol. 28, pp. 333–335; Olayiwola Abegunrin and Olusoji Akomlafe, “Nigeria and the Southern African Liberation Struggle,” in Abegunrin and Akomlafe eds., Nigeria in Global Politics (New York: Nova Science, 2006), p. 115; and Wright, The Destruction of a Nation, p. 72.

122. 

Memorandum of Conversation, December 19, 1975, 6 p.m., FRUS 1969–1976, Vol. 28, pp. 399–403.

123. 

Wright, The Destruction of a Nation, p. 72.

124. 

Memorandum of Conversation, December 18, 1975, FRUS 1969–1976, Vol. 28, p. 391.

125. 

Westad, The Global Cold War, p. 237.

126. 

Wright, The Destruction of a Nation, p. 76.

127. 

For Britain, see Jacqueline Audrey Kalley, Southern African Political History: A Chronology of Key Political Events from Independence to Mid-1997 (Westport, Conn.: Greenwood, 1999), p. 2. For France and Canada, see Wright, The Destruction of a Nation, p. 76; and “Angola: Recognition, Not Control,” Time, March 1, 1976, p. 37, http://content.time.com/time/magazine/article/0,9171,918111,00.html.

128. 

Letter from French President Giscard d'Estaing to President Ford, February 12, 1976, FRUS 1969–1976, Vol. 28, p. 437.

129. 

Memorandum from the President's Assistant for National Security Affairs (Scowcroft) to President Ford, undated, FRUS 1969–1976, Vol. 28, p. 474; and Note from the White House Staff Secretary (Connor) to President Ford, May 11, 1976, FRUS 1969–1976, Vol. 28, p. 476.

130. 

Wright, The Destruction of a Nation, p. 76.

131. 

Office of the Historian, “Angola,” A Guide to the United States' History of Recognition, Diplomatic, and Consular Relations, by Country, since 1776 (Washington, D.C.: U.S. Department of State, n.d.), https://history.state.gov/countries/angola.

132. 

“Angolan Ceasefire Signed,” BBC News, April 4, 2002, http://news.bbc.co.uk/2/hi/africa/1910419.stm.

133. 

The United States communicated widely its intelligence on Soviet and Cuban involvement in Angola. See Telegram from the Department of State to Certain African Diplomatic Posts, January 3, 1976, 1:47 a.m., FRUS 1969–1976, Vol. 28, p. 421.

134. 

Minutes of a National Security Meeting, June 27, 1975, 2:30–3:20 p.m., p. 266.

135. 

The first Cambodian civil war included two transfers of sovereign authority, one by coup (1970) and one by conquest (1975). I only discuss the details of the coup inasmuch as it determined the parties to the conflict. A number of governments continued to recognize Sihanouk's ousted government-in-exile, but I do not consider those recognitions to include recognition of the Khmer Rouge after its alliance with Sihanouk.

136. 

Ratliff, “UN Representation Disputes,” p. 1244.

137. 

Wilfred P. Deac, Road to the Killing Fields: The Cambodian War of 1970–1975 (College Station: Texas A&M University Press, 1997), pp. 53, 60.

138. 

Memorandum from Richard Kennedy of the National Security Council Staff to the President's Assistant for National Security Affairs (Kissinger), August 6, 1973, FRUS 1969–1976: Vol. 10: Vietnam, January 1973–July 1975 (Washington, D.C.: GPO, 2010), pp. 394–395.

139. 

Deac, Road to the Killing Fields, pp. 147, 203.

140. 

For example, see Transcript of Telephone Conversation between President Nixon and the President's Assistant for National Security Affairs (Kissinger), August 12, 1973, 2:30 p.m., FRUS 1969–1976, Vol. 10, pp. 420–421.

141. 

Ratliff, “UN Representation Disputes,” pp. 1245–1247.

142. 

“Cambodia—Resolution for Representation of Royal Government Rejected by UN General Assembly,” Keesing's Contemporary Archives, Vol. 21 (1975), p. 27149.

143. 

UN General Assembly, “Second Report of the Credentials Committee (A/9779/Add.1)” (New York: UN, December 13, 1974).

144. 

“Cambodia—The Siege of Phnom Penh,” Keesing's Contemporary Archives, Vol. 21 (1975), p. 27150.

145. 

“Cambodia—Foreign Recognition of Royal Government,” Keesing's Contemporary Archives, Vol. 21 (1975), p. 27152.

146. 

“Cambodia—The Siege of Phnom Penh,” pp. 27151–27152.

147. 

“Cambodia—Foreign Recognition of Royal Government,” p. 27152; and Peterson, Recognition of Governments, pp. 43–44. Colin Warbrick disagrees with this interpretation of French behavior. See Warbrick, “Kampuchea: Representation and Recognition,” International Comparative Law Quarterly, Vol. 30, No. 1 (January 1981), p. 234, doi:10.1093/iclqaj/30.1.234.

148. 

“Cambodia—Diplomatic Relations,” Keesing's Contemporary Archives, Vol. 21 (1975), p. 27471.

149. 

Warbrick, “Kampuchea,” p. 234.

150. 

UN General Assembly, “Report of the Credentials Committee (A/10270)” (New York: UN, October 1, 1975); and Mats Berdal and Michael Leifer, “Cambodia,” in James Mayall, ed., The New Interventionism 1991–1994: United Nations Experience in Cambodia, Former Yugoslavia, and Somalia (Cambridge: Cambridge University Press, 1996), p. 28.

151. 

The struggle between the United States and France for Chad's oil began in 1960. French oil firms, primarily Elf, were entitled to prospect in southern Chad immediately following its independence, but Chad's first president, Francis Tombalbaye, also granted exploration rights to U.S. firms. The relative success of U.S. firms caused a rift between Tombalbaye and the French government. After the coup in 1975, the French pressured Félix Malloum to rescind the U.S. concession. Chad's oil and other mineral potential went unverified until 1988, when a legal framework for exploration was put into place under Hissène Habré. Humphreys finds that although Habré “is reported to have received US support in exchange for his support of US oil corporations,” rebel leader Idriss Déby “was offered military support from France's Elf if he would overthrow the Habré régime and give France a stake in the southern oil fields.” Under Déby, who became Chad's president in 1990, France extracted a 20 percent stake in the southern oil fields from Exxon-Shell. See Humphreys, “Natural Resources, Conflict, and Conflict Resolution,” p. 509; and Burr and Collins, Africa's Thirty Years' War, p. 277.

152. 

Mario J. Azevedo and Emmanuel U. Nnadozie, Chad: A Nation in Search of Its Future (Boulder, Colo.: Westview, 1998), p. 48.

153. 

Ibid.

154. 

Geoff Simons, Libya and the West: From Independence to Lockerbie (Oxford: Centre for Libyan Studies, 2003), p. 55.

155. 

Azevedo and Nnadozie, Chad, pp. 48–49.

156. 

Simons, Libya and the West, p. 56.

157. 

“Chad,” in Colin Legum, ed., Africa Contemporary Record, Vol. 4: 1971–72 (New York: Africana, 1972), pp. B487, B490.

158. 

Azevedo and Nnadozie, Chad, p. 49; and Simons, Libya and the West, p. 56.

159. 

Foltz, “Chad's Third Republic,” p. 1.

160. 

Sam C. Nolutshungu, Limits of Anarchy: Intervention and State Formation in Chad (Charlottesville: University Press of Virginia, 1996), pp. 95–96.

161. 

Ibid., p. 107; and Alex Rondos, “Why Chad?” CSIS Africa Notes, August 31, 1983, p. 3.

162. 

Azevedo and Nnadozie, Chad, p. 53.

163. 

Samuel Decalo, “Chad: The Roots of Centre-Periphery Strife,” African Affairs, October 1, 1980, pp. 204–205.

164. 

Foltz, “Chad's Third Republic,” p. 2; and Rondos, “Why Chad?” p. 4.

165. 

Rondos, “Why Chad?” p. 4.

166. 

Robert Buijtenhuijs, “The Chadian Tubu: Contemporary Nomads Who Conquered a State,” Africa, Vol. 71, No. 1 (February 2001), p. 151.

167. 

UN General Assembly, “Second Report of the Credentials Committee (A/34/500/Add.1)” (New York: UN, November 14, 1979).

168. 

Azevedo and Nnadozie, Chad, p. 54.

169. 

Rondos, “Why Chad?” p. 5.

170. 

Ibid., p. 7.

171. 

Foltz, “Chad's Third Republic,” p. 2.

172. 

Decalo, “Chad,” p. 91; and Burr and Collins, Africa's Thirty Years' War, p. 157.

173. 

Rondos, “Why Chad?” p. 8.

174. 

Foltz, “Chad's Third Republic,” p. 2.

175. 

Ibid.

176. 

“Chad—Military Campaign of FAN,” Keesing's Contemporary Archives, Vol. 28 (1982), p. 31679.

177. 

Ibid.

178. 

UN General Assembly, “Report of the Credentials Committee (A/37/543)” (New York: UN, October 26, 1982).

179. 

Simons, Libya and the West, p. 58.

180. 

Foltz, “Chad's Third Republic,” p. 2.

181. 

Burr and Collins, Africa's Thirty Years' War, p. 262.

182. 

“Chad,” in Colin Legum, ed., Africa Contemporary Record, Vol. 23: 1990–92 (New York: Africana, 1998), p. B201.

183. 

Burr and Collins, Africa's Thirty Years' War, pp. 263–264.

184. 

“Chad,” Vol. 23, p. B202.

185. 

UN General Assembly, “Report of the Credentials Committee (A/46/563)” (New York: UN, October 11, 1991).

186. 

Burr and Collins, Africa's Thirty Years' War, pp. 131–33, 152–156.

187. 

Unless otherwise noted, the remaining cases are all defined as conflicts in the data set (N = 61), less the five already discussed in detail (Angola, Cambodia I, Chad I, Chad II, and Libya).

188. 

The exceptions are: Tajikistan, where Russia extended recognition to Rakhmanov two weeks before he seized the capital; and Lebanon II, where the recognized coalition controlled half the capital at the time of recognition.

189. 

I report the median because of two outlier cases (Lebanon II, −356 days; and Taliban/Afghanistan, 240 days). The mean is −4 days. Lebanon II was a case of partial capital city control for 356 days.

190. 

The Taliban regime in Afghanistan was bilaterally recognized by three states (Pakistan, Saudi Arabia, and the United Arab Emirates). Pakistan campaigned for UN recognition on the basis of the Taliban's control of “almost two-thirds of the country, including the capital, Kabul.” See UN Department of Public Information, Yearbook of the United Nations, Vol. 51 (New York: UN, 1997), p. 252. This argument was ignored by other UN member states, though the Taliban's absolute level of control (territory and capital) is consistent with other cases of positive recognition.

191. 

For a discussion of these maneuvers, see Griffin, “Accrediting Democracies,” pp. 730–731 n. 17.

192. 

Afghanistan is coded as denying recognition, because it suspended diplomatic contact (initiated during the prior post-Communist government) and the Burhanuddin Rabbani government provided military support to anti-Rakhmanov forces based in Afghanistan. See Barnett R. Rubin, “The Fragmentation of Tajikistan,” Survival, Vol. 35, No. 4 (1993) pp. 82–86, doi.org/10.1080/00396339308442712.

193. 

I code a rebel campaign as prioritizing the capital if there is no apparent attempt by a rebel group to establish a formal hold over territory other than the capital prior to gaining sovereign authority. For an example, see Laurent Kabila's sprint across DRC in 1996, as documented in Prunier, Africa's World War, pp. 133–134.

194. 

The total 61 cases include 18 contested capitals, with 12 suffering significant destruction or displacement of residents. I code a capital as experiencing “significant destruction” if sources report more than 50 percent of infrastructure being damaged or looted, or if they report substantial use of heavy artillery. I code the residents of a capital as experiencing “significant displacement” if sources report the city as “empty” or “abandoned” following the conflict.

195. 

Collier et al., Breaking the Conflict Trap, p. 15.

196. 

Human Rights Watch, Blood-Stained Hands, pp. 22–100.

197. 

UN Security Council, “Report of the Secretary-General on the Situation in the Republic of Congo (S/1997/814),” pp. 7–8.

198. 

Duyvesten, Clausewitz and African War, p. 113; Drysdale, Whatever Happened to Somalia? p. x; and Simons, Networks of Dissolution, p. 91.

199. 

The median conflict duration is 4.9 years after first bilateral recognition, and the mean is 7.5 years.

200. 

Carl von Clausewitz argues that the capital is “the center of gravity” during civil war. See Clausewitz, On War, Michael Howard and Peter Paret, trans. and eds. (Princeton, N.J.: Princeton University Press, 1989), p. 596.

201. 

For extensive reviews of the standards used until 1945, see Peterson, Recognition of Governments; and Lauterpacht, Recognition in International Law, Vol. 3, pp. 98–140.

202. 

For a discussion of willingness to honor international obligations, see Lauterpacht, Recognition in International Law, p. 113. For a discussion of popular endorsements, including referendums, see ibid., pp. 125–130. For imposing stipulations generally, see Peterson, “Political Use of Recognition,” pp. 337–341.

203. 

For a detailed examination of Article 2 of the General Treaty of Peace and Amity in 1923, which forbade recognition of new governments emerging from coups or revolutions in the signatory Central American countries, see Charles Cheney Hyde, International Law, Chiefly as Interpreted and Applied by the United States, Vol. 1, 2nd ed. (Boston: Little, Brown, 1947), pp. 173–174. For developments in Africa, see Claire Felter, “Africa's ‘Leaders for Life’ Syndrome” (New York: Council on Foreign Relations, December 28, 2017), https://www.cfr.org/backgrounder/africas-leaders-life-syndrome.

204. 

Ratliff, “UN Representation Disputes,” pp. 1233–1256.