Establishing a practical test for the end of non-international armed conflict

Abstract

When do non-international armed conflicts (NIACs) end? Determining the existence of a NIAC requires a detailed, fact-intensive inquiry. Since the International Criminal Tribunal for the Former Yugoslavia's seminal decision in the Tadić case, international courts and tribunals have evaluated the existence of NIACs under the Tadić test's two-pronged inquiry into intensity and organization. Although that decision also pronounced that international humanitarian law (IHL) continues to apply until a “peaceful settlement is achieved,” neither international tribunals nor scholars have articulated a comparably widely accepted and well-developed test for determining the end of NIACs.

At the same time – and especially since 9/11 – States have increasingly relied on IHL to meet the threat posed by non-State actors, broadening the scope of conflict-related liabilities in armed conflict without conferring the privileges or immunities otherwise inherent in IHL. This one-sided approach twists the purpose of IHL and places members of organized armed groups into legal black holes without temporal limitations, as States resist the termination of “armed conflict” irrespective of the continuing intensity of violence or the level of organization of non-State actors. Ultimately, the current approach gives States broad discretion without appropriate safeguards, which undermines the proper application of human rights and humanitarian principles within conflict and prevents the establishment of a sustainable peace.

This paper argues that the most appropriate test for ascertaining the end of a NIAC is one that combines objective consideration of the diminution of organized and intense hostilities below the Tadić threshold with the likelihood that hostilities will not again rise above that threshold. It thus draws from but does not fully endorse the preferred approach of the International Committee of the Red Cross (ICRC), which focuses on the lasting termination of hostilities,¹ while abjuring a general temporal limitation. Although commendable for its effort to avoid the legal uncertainty that attends the revolving conflict classification problem, the ICRC's approach unfortunately tends to encode existing uncertainty surrounding the termination of NIACs and to indistinctly prolong the application of IHL to erstwhile conflict situations. In contrast, the authors suggest that a test for the end of NIAC based on a specific period (five months following the diminution of organized and intense hostilities below the Tadić threshold), subject to an evaluation of the risk that those hostilities may resume, better balances certainty of legal application with the promotion of a return to peace. The authors will employ the facts of diverse case studies, including the FARC in Colombia, the LTTE in Sri Lanka, numerous armed groups in Mali, and the United States with Al-Qaeda, to build a legal standard that courts can use to determine IHL's continuing applicability to an erstwhile armed conflict.

Introduction

How do non-international armed conflicts (NIACs) end? Although international humanitarian law (IHL) is inherently temporally bounded,² there is no controlling test for assessing whether or when a given NIAC has terminated³ beyond the indistinct pronouncement of the International Criminal Tribunal for the former Yugoslavia (ICTY) that IHL continues to apply to a NIAC until a “peaceful settlement is achieved”.⁴ As recently as 2010, the International Law Association found that the end of the temporal scope of application of IHL is a “complicated issue … in need of thorough research”.⁵ And uncertainty concerning when NIACs end, as a matter of law, continues to persist⁶ despite the imperative of delineating when IHL applies – when the regulation of killing and deprivation of liberty, for example, is relaxed – and when it does not apply – when killing and deprivation of liberty are strictly limited.⁷

In lieu of an authoritative test for determining the end of a NIAC, scholars have proposed four competing tests for the termination of NIACs: the “reverse Tadić test”, which emphasizes whether an ongoing conflict continues to satisfy the two-pronged test of intensity and organization developed by the ICTY;⁸ the “no-more-combat-measures test”, which analogizes to the termination of international armed conflicts (IACs) and focuses on the persistence (or not) of military activities in a given armed conflict; the “peaceful settlement test”, which draws from Tadić but additionally asks whether a political solution to an armed conflict has been achieved; and the “no-reasonable-risk-of-resumption test”, which employs a variety of factors to assess whether hostilities are likely to resume.⁹ In their own ways, each of these formulations endeavours to articulate a legal test that parameterizes “peace” as the descriptive absence of armed conflict – but because NIACs exist across a factually diverse spectrum, none of the proposed tests satisfyingly describes all instances in which warfare has concluded, and none of them has secured authoritative stature in international law. The continuing absence of a definitive test for determining the end of NIACs undermines the efficacy of IHL, inhibits the (re-)establishment of peace, prolongs conflict-related civilian vulnerability, and challenges the international rule of law.

This article endeavours to correct this deficiency of IHL by formulating a fact-based test for determining when NIACs and the application of IHL terminate as a matter of law. It surveys the four competing tests and argues that the reverse Tadić test, subject to a durational requirement of five months and objective evaluation of the risk that hostilities will resume, would more effectively describe the end of NIACs in fact. Under our standard, the end of a NIAC would be determined on the basis of the diminution of intensive, organized armed violence; the length of time since hostilities fell below the Tadić threshold; and the quality of the conflict's pacification. This standard would cabin States’ discretion to claim the continued persistence of a NIAC despite contrary, observable factors. It would entail the resumption of the lex generalis, international human rights law, even when sporadic, isolated organized armed violence or disorganized armed violence continues – as would be the case if the situation never crossed the NIAC threshold in the first place. In such circumstances, that sporadic organized or disorganized violence would, appropriately, be primarily regulated by domestic criminal law.

This article begins by describing the consequences of indeterminate NIACs. It then surveys the four competing tests for the end of a NIAC, and explores the limitations of those tests in the context of four recent NIACs. Finally the article describes our suggested expansion of the “factual and lasting pacification of NIAC” standard and applies it to the previously described case studies.

The consequences of indefinite NIACs

Just as it may be dangerous to belligerent parties to prematurely treat a NIAC as ended, it is dangerous to indefinitely prolong the legally exceptional existence of an armed conflict. Doing so exposes civilians to the vulnerabilities inherent in a legal regime that accepts¹⁰ lethal force as a first resort, that admits the incidental killing of civilians,¹¹ and that legitimizes deprivation of liberty with reduced substantive and procedural protections. As the United States’ prosecution of the so-called Global War on Terror illustrates, States may embrace a NIAC framework and exploit its comparatively liberal use-of-force regime to broadly treat antagonists as detainable or targetable enemies, contributing to both increased civilian casualties and otherwise impermissibly low evidentiary standards in legal proceedings supposedly related to the NIAC.

Indefinite NIACs may also benefit non-State actors who seek to wear down States through prolonged, armed violence that would be prohibited in ordinary circumstances.¹² The present indeterminacy of the legal end of NIACs means that even armed groups which no longer satisfy the Tadić test's organizational requirement or which, while organized, prosecute hostilities that may no longer satisfy that test's intensity requirement may claim the implied authority for violence under IHL. In contrast, a fact-based test for the legal end of NIACs would more appropriately subject such groups or such low-intensity violence to ordinary criminal law, enforced by a functioning judiciary.

Indefinite war wreaks havoc on civilians within conflict zones. Among other vulnerabilities, they often become dependent on humanitarian aid, which in turn is dependent on the vagaries of the conflict parties. During the Sri Lankan Eelam Wars, much of the population in northern Sri Lanka lived in displaced persons’ camps, and the Sri Lankan government blocked both food aid and information regarding civilian suffering due to lack of food aid.¹³ It was often impossible to continue fishing on the northern coast, as it was occupied by soldiers. While conflict parties may rely on the emergent nature of hostilities to justify such transgressions, a fact-based test for the legal end to NIACs may allow for proper accountability for what, in peacetime, are gross human rights and criminal violations.

Existing tests for the end of NIACs

Nearly thirty years ago, the ICTY found that IHL continued to apply to NIACs until a “peaceful settlement is achieved”.¹⁴ Since then, courts have been reluctant to determine whether or when NIACs conclude, as a matter of IHL. During that period, scholars developed four competing tests to ascertain whether or when a peaceful settlement of a NIAC has been achieved – none of which has become widely accepted. Fundamentally, none of the four theories addresses the “messiness” of NIACs.

Reverse Tadić

The reverse Tadić test contemplates that a NIAC ends when one of its constituent conditions, intensity of violence or organization of a party, falls below the threshold required to establish hostilities.¹⁵ While elegant in its symmetry, practical application of this test is challenged by the tendency for intensity of violence in NIACs to ebb and flow. Likewise, NIACs involving multiple parties or situations in which overlapping NIACs occur simultaneously within a single conflict zone or across multiple conflict zones (“complex NIACs”) may be dynamic, consisting of rotating or even morphing groups that fluctuate in their level of organization or participation in hostilities over time. The result of embracing the reverse Tadić test, particularly without a temporal standard or evaluation of the risk of renewed hostilities, would be considerable legal fluctuations, creating operational confusion among the parties and impeding efforts to impose accountability for war crimes. As the International Committee of the Red Cross (ICRC) has admonished, it is preferable to avoid a “revolving door classification” of a conflict that generates more confusion than clarity.¹⁶

No more combat measures

The no-more-combat-measures test borrows from the practice in IACs of tying the end of hostilities to the end of military operations.¹⁷ Several complications present themselves when applying this test to actual NIACs. First, in complex NIACs, it can be difficult to stop the clock on a conflict between any two parties while it continues with respect to other parties. Relatedly, outright military-style action may end for a period as a group refits, rearms or reorganizes, only to resume slightly later, once again raising the revolving-door problem. On the other hand, waiting for an (often politically influenced) announcement of the close of military operations to apply the no-more-combat-measures test “may prolong and overextend the application of [IHL], which permits forms of violence that would otherwise [under ordinary domestic law] be unlawful and impermissible according to general international law”,¹⁸ thereby undermining the objective of restoring peace.

Peaceful settlement

The peaceful settlement test, which also leans heavily on limited objective measures more practicable in IACs, requires a peaceful settlement among the parties to a NIAC. It has the virtue of being simple, and easily understood and applied by a variety of interested actors and observers.¹⁹ However, this theory discounts the frequent refusal of States to recognize the legitimacy of non-State actors or even enter into negotiations with their non-State adversaries. Thus, this test might preclude the termination of NIACs where States (or non-State organized armed groups) simply refuse to discuss the termination of the conflict. The requirement for a peaceful settlement may also be wildly unrealistic for conflicts involving multiple non-State actors with different capabilities and objectives.

No reasonable risk of resumption

The no-reasonable-risk-of-resumption test is the only one to incorporate, in part, a temporal condition. This test recognizes that quieted NIACs must be evaluated as to whether it would be “unreasonable” to believe that hostilities might resume. Such evaluation may include assessment of the duration of time without hostilities; statements made by parties regarding their intention to end violence; actual steps taken (including disarmament) to enforce peace; and the existence of a peace agreement. Importantly, this theory would not rule out the end of a NIAC on the basis of “isolated or sporadic” acts of violence, thereby implementing the symmetry of the reverse Tadić test. The “reasonableness” standard has been criticized as vague, but in fact echoes objective reasonableness standards applied in both common law and international criminal law. The absence of definitive criteria, however, may lead different assessors to weigh the same factors differently, arriving at disparate legal conclusions.

The ICRC version of this standard²⁰ involves a cessation of hostilities and a fact-based analysis of indicators relevant to whether the conflict is likely to reignite. Unfortunately, it injects additional uncertainty into determining whether or when a NIAC has ended by suggesting conflict-specific historical assessments of temporary quiescence to determine whether resumption of hostilities is reasonably likely.²¹ We agree that none of these four theories on their own provide the proper guidance needed to safely and accurately pronounce the end of a NIAC. Therefore, following the case studies below, we draw upon elements of all four theories for a proposed three-step analysis to determine the end of a NIAC.

Case studies

Just as the Tadić test is based on a highly fact-specific analysis, the starting point for ascertaining the end of a NIAC must be through similar factual examination of different situations and their common elements. In this section, we review four situations, representing four different areas of the world, that have been categorized at various times as NIACs. We analyze each of these situations through the lenses of the four competing end-of-NIAC tests in order to illuminate the limitations of those tests. The proposed three-step analysis draws heavily from the political and humanitarian challenges faced in these case studies, and the limitations they reveal in the competing end-of-NIAC tests.

Colombia and the FARC

In 2016, the fifty-two-year NIAC between the Revolutionary Armed Forces of Colombia (Fuerzas Armadas Revolucionarias de Colombia, FARC) and the government of Colombia ended in an unusually satisfying manner: a peace agreement between the parties. The FARC–Colombia NIAC arose in May 1964 out of violence following an attack by Colombia's armed forces against a rural, communist militia that had declared its own republic in defiance of the government.²² According to Colombian government figures, the armed conflict killed at least 220,000 people over five decades.²³ It displaced between 5.7 million²⁴ and as many as 7 million Colombians.²⁵ The FARC grew from just forty-eight men in 1964²⁶ to a force that, at its peak, boasted 20,000 fighters²⁷ and controlled 42,000 square miles of territory.²⁸ The FARC's armed forces were organized hierarchically into seven blocs, each of which was responsible for at least five fronts and answered to the FARC's Secretariat and Central General Staff.²⁹ When it finally entered into a peace agreement with Colombia, the FARC had 7,000 fighters and 2,800 militia, dispersed over 22% of Colombian municipalities³⁰ – many of which it administered as the de facto public authority.³¹

Following years of secret and complex negotiations, the FARC–Colombia peace agreement provided no mere paper end to hostilities. Combat between the FARC and the government ended, and nearly 14,000 FARC members – including fighters, militia and prisoners – surrendered their weapons to the United Nations (UN) Mission in Colombia,³² demobilized,³³ and matriculated into a government-sponsored reintegration programme.³⁴ The agreement ushered the conversion of the FARC into a political party, with temporarily guaranteed representation in the Colombian legislature.³⁵ The agreement also enabled the Colombian government to reassert its authority over large swathes of territory.³⁶

Still, the peace settlement between the FARC and the government of Colombia has not been perfect. The agreement encompassed only the FARC and the Colombian government, leaving unaddressed the separate and ongoing NIAC between the government and the National Liberation Army. Reportedly, FARC leaders sought to maintain the organization's chain of command throughout the demobilization process in order to ensure the group's post-conflict political future;³⁷ and some dissenters within the FARC's ranks refused to abide by the agreement, forming a new organized armed group.³⁸

The 2016 termination of the FARC's NIAC with Colombia may be a paradigmatic example of the end of a NIAC. Seven years on, it seems clear that a “peaceful settlement [has been] achieved” – at least between the government and the FARC itself. The parties negotiated a political solution to the armed conflict, which ended in fact hostilities between the parties. Thus, this case study provides good – if arguably rare – evidence of the viability of the peaceful settlement test for the end of a NIAC. Likewise, because the intensity of the armed conflict fell below the threshold necessary to satisfy the Tadić test even before the peace agreement and military activities ended thereafter, this case study suggests the viability of both the reverse Tadić test and the no-more-combat-measures test.

Nevertheless, closer examination of the situation reveals dissatisfying qualities of each of the competing tests. Significantly, hostilities between the FARC and Colombia did not cease in November 2016, when the agreement became operative. The FARC and Colombia entered into an indefinite ceasefire nearly six months before the formal peace agreement was finalized.³⁹ That ceasefire not only followed a series of temporary unilateral ceasefires, but also came in a period of relative quiescence in the conflict as measured by the number of deaths due to armed violence between the parties. According to the Uppsala Conflict Data Program, there were fewer than 400 deaths as a result of armed violence between the FARC and Colombia each year from 2007 on.⁴⁰ In the five years before the peace agreement was finalized, armed violence between the two parties killed on average only 145.8 individuals annually.⁴¹ This measure alone suggests a diminution in intensity of the armed conflict such that it may have no longer satisfied the Tadić test well before the peace agreement entered into force.

The no-more-combat-measures test, however, points to a markedly different conclusion. Both the FARC and the government of Colombia continued to engage in sporadic combat activities before and after the 25 June 2016 bilateral and lasting ceasefire. Indeed, the FARC may have been responsible for four additional attacks against Colombian governmental or private targets between 25 June and 9 December 2016.⁴² Thus, the no-more-combat-measures test would indicate that the FARC–Colombia NIAC continued for at least several weeks beyond the date on which the peace agreement took effect.⁴³

While the conclusion of the formal peace agreement between the FARC and Colombia provides a clear statement that the parties intended to end the armed conflict, at the moment it entered into force the fact of the agreement itself necessarily could say little about the real risk of the conflict's resumption. The durability of the peace agreement is only evident in retrospect in light of corroborating facts like the massive demobilization of FARC fighters and the surrender of their weapons, which was not completed until 27 June 2017.⁴⁴ Thus, while the peaceful settlement test might require the inapplicability of IHL as of November 2017, the no-real-risk-of-resumption test might require continued application of IHL until the FARC's demobilization was complete some seven months later.

Even in the case of the FARC–Colombia armed conflict, then, the four competing tests prove unsatisfactory, pointing to four disparate conclusions about the date on which IHL ceased to apply to that armed conflict.

The Sri Lankan civil war

The Sri Lankan civil war began in July 1983 after riots in Colombo targeting the Tamil community. Much of the thirty-five years since independence until that point had seen deepening hostility between the Sinhalese majority and the Tamil minority, which had begun in large part due to the British government's pre-independence promotion of and general favour towards the Tamil population.⁴⁵

Over three decades after independence, the Sinhala population gained power and enacted policies that resulted in disenfranchisement of the Tamils. Because the Tamil population was concentrated primarily in the northern and eastern parts of the island, a number of Tamil groups coalesced around the demand for an independent Tamil State there. The Liberation Tigers of Tamil Eelam (LTTE) eventually defeated the other separatist groups and came to represent the Tamil independence movement.⁴⁶ The LTTE was organized into regiments with a clear hierarchy and grew to include political and financial wings.⁴⁷ While the independence movement was premised on the concept of self-determination,⁴⁸ the LTTE's tactics broadened beyond military targeting to attacks upon the civilian population and the use of child soldiers, human shields and suicide bombings.⁴⁹

The predominately Sinhalese Sri Lankan government also initiated violent attacks against the Tamil population (beyond clashes with the LTTE) during the civil war. While the Sri Lankan Defence Force was weak and disorganized compared to the much smaller LTTE (one factor that contributed to the length of the conflict),⁵⁰ the central government detained political dissidents without trial and used torture and extrajudicial killings with impunity.⁵¹ The government also denied civilian populations access to necessities including food and clean water.⁵² While estimates vary, at the height of its power, the LTTE controlled large areas of northern and eastern Sri Lanka.⁵³

The decades of conflict in Sri Lanka are generally broken into four “Eelam Wars”.⁵⁴ The first Eelam War dated from the initial riots in 1983 until the 1987 Indo-Sri Lankan Peace Accord between Sri Lanka and the Indian government, which had intervened on behalf of Tamils in the north. Over the next three years, the LTTE refused to disarm and killed more than 1,000 Indian soldiers.⁵⁵ India withdrew from the conflict in 1990, and the Second Eelam War dated from the June 1990 collapse of peace talks and the LTTE massacre of nearly 800 police who had surrendered in the north, until a ceasefire in January 1995.⁵⁶ That ceasefire was broken in April 1995 by the LTTE, and the ensuing Third Eelam War lasted until about February 2002.⁵⁷ In December 2001, Norway mediated peace talks between the Sri Lankan government and the LTTE that culminated in what was meant to be a permanent ceasefire agreement (CFA) in February 2002. The period of peace talks was marked by the lifting of an economic embargo on LTTE-controlled territory⁵⁸ and the first prisoner exchanges since the set of conflicts began in 1983. The CFA lasted for four years, but the country descended into violence again in July 2006, the start of the Fourth Eelam War, when the LTTE cut off a vital water supply to government-controlled areas. After a year of fighting, the government reclaimed control of territory in the east, and in 2008 it formally withdrew from the 2006 CFA. The formal conflict ended in May 2009, with the Sri Lankan forces boxing in and eliminating a significant portion of the LTTE leadership, including its leader.⁵⁹

The twenty-six-year Sri Lankan conflict was punctuated by periods of greater or lesser quiescence that delineate the four Eelam Wars. There were five turbulent years between the First and Second Eelam Wars, only 100 days between the Second and Third, and about four years between the Third and Fourth. While the first period of “peace” never saw a cessation of violence, and 100 days of tenuous peace is not likely to inspire confidence in ending a NIAC without other strong indicators such as full disarmament, the last period of “peace” may arguably be considered to have triggered the end of that NIAC.

The 2012 violence in Mali

Between January and April 2012, violence erupted in northern Mali in what was has been referred to as the “Azawadi” (independence) movement among several armed groups, against the Malian government.⁶⁰ Among them, the minority Tuareg population in Mali had long complained of marginalization and discrimination.⁶¹ The initial insurgency was led by the National Movement for the Liberation of Azawad (Mouvement National pour la Libération de l'Azawad, MNLA), which incorporated Tuareg, Arab and Songhai fighters and sought an independent State in northern Mali.⁶² It was joined by other groups, including Ansar Dine, the Movement for Oneness and Jihad in West Africa (Mouvement pour l'Unicité et le Jihad en Afrique de l'Ouest, MUJAO) and Al-Qaeda in the Islamic Maghreb (AQIM),⁶³ which had different ends, and some of which clashed with each other.⁶⁴ The MNLA, Ansar Dine and MUJAO jostled for control of the cities of Kidal, Gao and Timbuktu; the MNLA was eventually pushed out of all three.⁶⁵ Importantly, following a coup d’état, the Malian Armed Forces (Forces Armées Maliennes, FAMa) largely deserted Timbuktu by the end of March 2012, and it was easily taken over by the rebel groups.⁶⁶ The disparate approaches of the groups combined with the desertion of the FAMa have led to the argument that the violence within Mali between January and April 2012 did not rise to the level of a NIAC at all⁶⁷ – a position bolstered by legal scholarship noting that the minimum period found to satisfy the protracted element of an armed conflict is five months.⁶⁸

Assuming the existence of a NIAC in 2012, the situation in Mali further demonstrates the insufficiency of the four competing tests. Under any single theory, any existing hostilities in Mali in early 2012 had arguably ended with the takeover and installation of administration in northern Mali by non-State actors by April 2012.⁶⁹ Under the reverse Tadić theory, the FAMa was sufficiently disorganized and disengaged to diminish organized hostilities below the Tadić threshold by April 2012. Therefore, the application of IHL would have ceased because “the conditions that triggered its application” no longer existed.⁷⁰ Similarly, the FAMa's collapse and the absence of hostilities among other groups following the establishment of non-State administration of Kidal, Gao and Timbuktu⁷¹ may have constituted a “general close of military operations” under the no-more-combat-measures theory, marking the termination of the NIAC.

A cessation of hostilities “without real risk of resumption” constitutes the end of a NIAC under the no-reasonable-risk-of-resumption theory,⁷² despite “minor isolated or sporadic acts of violence”, which could characterize the events in Mali between April 2012 and January 2013. While the Diakonia International Humanitarian Law Centre in Mali has determined that “from 2012 to 2015, Mali's regular army was engaged in a NIAC against [specifically] the MNLA”, the Centre also notes that “the description of [acts by other non-State actors] by third parties sometimes reflects more their interest in portraying these groups as organized entities, rather than a real degree of organization”.⁷³

Finally, a peaceful settlement arguably existed among the formerly warring parties by the end of March 2012, as reflected in the lack of violence surrounding the administration of Timbuktu and other northern cities, despite transfers of power from the MNLA to Ansar Dine and AQIM. Under the peaceful settlement theory, “what counts is the pacification of the situation, not the disappearance of the criteria” – a “peaceful settlement is to be understood in its material sense and not in the formal sense of a peace treaty or another agreement of the same kind”.⁷⁴ Although some parties objected to the replacement of the sitting government and the implementation of Islamic law by the new groups, it does not legally follow that such governance was not largely peaceful, or did not constitute a peaceful settlement of any previous hostilities prior to April 2012. Similarly to the period of “peace” between the Third and Fourth Eelam Wars in Sri Lanka, the administration of northern Mali by non-State actors after April 2012 appears to have been relatively stable, with the non-State actors described by international media as “unshakable in their stronghold”.⁷⁵ The “real risk of resumption” is perhaps the “X-factor” here; in April 2012, it seemed unlikely that the FAMa would have attempted to retake Timbuktu or the other northern cities on its own.⁷⁶ In fact, the Malian government asked for international assistance and believed until January 2013 France's statement that “it wouldn't intervene”.⁷⁷ The French intervention in January 2013 was therefore not foreseeable such that the characterization of the Malian conflict as a NIAC throughout was justified.⁷⁸ In fact, since without the French intervention, northern Mali may have remained separate from the administration of southern Mali, it can be argued that a NIAC ended in April 2012 and another armed conflict began in January 2013, with the French and Malian governments on one side against disparate non-State actors on the other.

Complicating the calculus is the fact that following Operation Serval in January 2013, violent clashes restarted in northern Mali that have continued to wax and wane through to the present, notwithstanding a number of peace deals and ceasefires between 2013 and 2015. The 2015 peace deal signed by multiple non-State groups, offering them a measure of autonomy in the north, slowed the violence from the minority components, but did not ultimately extinguish attacks sponsored by AQIM and other militant groups in the north. In assessing the conflict (or multiple conflicts) post-2013, it may be important to note that Operation Barkhane, the French intervention that succeeded Operation Serval, began in August 2014 and lasted through to November 2022.

The United States versus Al-Qaeda

The US–Al-Qaeda NIAC casts the deficiencies of the four competing tests for the end of a NIAC in stark relief. Applied to identical facts, those tests point to dramatically divergent legal conclusions as to the continuing existence of that NIAC. It also underscores the difficulty of assessing whether and when complex NIACs terminate. In part, this difficulty arises from the fact that the US–Al-Qaeda NIAC is complex both because it began in the context of overlapping armed conflicts in a single geographical space⁷⁹ – Afghanistan – and because it involved multiple parties. The complexity of this NIAC is only increased by the fact that Al-Qaeda's supposed co-belligerents⁸⁰ – “affiliates”, in US terminology – are dispersed globally. Thus, in assessing whether the US–Al-Qaeda NIAC may have ended, observers are challenged to distinguish between clashes with “core Al-Qaeda” and clashes with its affiliates, as well as whether any given organized armed group affiliated with Al-Qaeda is in law a co-belligerent of Al-Qaeda. Observers are likewise challenged by data limitations concerning whether Al-Qaeda fighters were present at or involved in clashes with US forces in Afghanistan during the Taliban's successful insurgency against the Islamic Republic of Afghanistan.

Notwithstanding those challenges, certain facts are ascertainable. First, under extant IHL, the US–Al-Qaeda NIAC began no earlier than the US invasion of Afghanistan on 7 October 2001.⁸¹ Second, since then, the United States and Al-Qaeda have not arrived at a political solution, although, on 21 September 2021, President Biden declared before the UN that “for the first time in 20 years … the United States [is] not at war”.⁸² Third, the intensity of violence between the United States and core Al-Qaeda likely fell below the threshold necessary to satisfy the Tadić test no later than December 2016. While the United States has continued to episodically and infrequently (so far as is publicly known) use military force against Al-Qaeda targets within⁸³ and without Afghanistan,⁸⁴ between December 2016 and June 2021, the United States did not even identify Al-Qaeda as one of the parties with which it was engaged in an armed conflict in Afghanistan in its semi-annual consolidated War Powers Resolution reports.⁸⁵ Outside of Afghanistan, Al-Qaeda reportedly has not been responsible for a terrorist attack anywhere since 2011,⁸⁶ has not conducted an attack in Europe since 2005⁸⁷ and has not attacked US territory since 11 September 2001.⁸⁸ Fourth, core Al-Qaeda may well be a spent force: according to the US intelligence community, “[c]onsistent U.S. and allied counterterrorism pressure has degraded the external attack capabilities of [Al-Qaeda]”,⁸⁹ even as “leadership and battlefield setbacks”⁹⁰ have caused Al-Qaeda to “[devolve] operational responsibility to regional affiliates”⁹¹ and to “[shift] away from centrally directed plotting”.⁹² Recent US intelligence community assessments emphasize the significance of the threat posed by Al-Qaeda's “affiliates” and “offshoots” (such as AQIM) in contradistinction to the relatively minor threat posed by Al-Qaeda itself.⁹³ According to the US Defense Intelligence Agency (DIA), core Al-Qaeda has “far fewer” than “200 members”⁹⁴ and potentially as few as a dozen legacy members of the organization in Afghanistan.⁹⁵ DIA further assesses that Al-Qaeda lacks the capability to attack the United States.⁹⁶ Finally, DIA assesses that Al-Qaeda's strength and capabilities have not improved despite the withdrawal of US forces from Afghanistan in August 2021 and the absence of any US military activity in Afghanistan targeting Al-Qaeda (except the strike killing the organization's emir in Kabul on 1 August 2022⁹⁷) since then.⁹⁸

Despite the years-long dearth of publicly identifiable clashes between the United States and Al-Qaeda, the United States maintains that, legally, its armed conflict against core Al-Qaeda has continued for more than twenty years.⁹⁹ As with the other NIACs surveyed in this article, the competing tests for the end of NIACs point to different legal conclusions as to whether the US–Al-Qaeda armed conflict continues – and when it may have concluded. For example, the reverse Tadić test suggests that the armed conflict ended no later than December 2016, when the violence fell below the level of intensity sufficient to constitute a NIAC in the first place. By then, Al-Qaeda had not been responsible for any publicly identifiable attacks for at least five years. Likewise, at that point, the United States ceased indicating that its forces were actively engaged in combat operations against Al-Qaeda, at least in War Powers Resolution reports pertaining to Afghanistan.

Similarly, the no-reasonable-risk-of-resumption test may indicate that the US–Al-Qaeda NIAC has long since ended. Assuming that the US intelligence community's assessments are correct, core Al-Qaeda continues to desire to attack the United States, which suggests that there is a risk that hostilities between the United States and Al-Qaeda will resume. However, again assuming the accuracy of the US intelligence community's assessments, Al-Qaeda lacks the capability to do so – an assessment that has not changed despite the Taliban's reassertion of control in Afghanistan and the withdrawal of US forces there nearly three years ago. In combination, the absence of US forces in Afghanistan and Al-Qaeda's failure to reconstitute itself despite their absence suggest that there may not be a reasonable risk that US–Al-Qaeda hostilities will reignite.¹⁰⁰ However, given Al-Qaeda's assessed continuing interest in attacking the United States, this conclusion is at best uncertain. The ambiguity of the no-reasonable-risk-of-resumption test in this context demonstrates its limitations as a test for determining the continuing applicability of IHL to complex NIACs because both operational necessities and the protective objective of IHL demand clarity and certainty.

On the other hand, the no-more-combat-measures test points to a conclusion that, consistent with the US position, the NIAC between the United States and Al-Qaeda continues. The United States continues to engage in military activity by detaining, subject to law-of-armed-conflict authorities, some members of Al-Qaeda at Guantanamo Bay.¹⁰¹ Likewise, the United States’ continuing, if episodic and infrequent, use of discrete military force against specific Al-Qaeda targets¹⁰² is an ongoing combat measure that perpetuates the US–Al-Qaeda armed conflict.¹⁰³

Finally, because the United States and Al-Qaeda have not reached a political solution, the peaceful settlement test indicates that the US–Al-Qaeda NIAC continues. Moreover, because it is extremely unlikely that the United States and Al-Qaeda will ever enter into a peace agreement, the peaceful settlement test points to a conclusion that the US–Al-Qaeda armed conflict will continue in perpetuity. But armed conflict – along with the application of IHL – is inherently a state of exception, and as a matter of logic, states of exception cannot be perpetual.

Proposed test

This paper argues that the most appropriate test for determining whether and when a NIAC has ceased requires an objective analysis of (i) whether hostilities have diminished below the Tadić threshold for the existence of a NIAC, (ii) the quality of pacification (including achievement of objectives consistent with international law), and (iii) the length of time since violence ceased and steps towards pacification began. The three elements laid out below combine objective factors from all four existing theories to more effectively analyze whether a NIAC has ended. Importantly, all three elements would have to be considered in order to achieve as conclusive an analysis as possible.

Element 1: Diminution of organized hostilities

Presently, it is impossible to legally conclude whether a NIAC has ended without a cessation of violence between two parties that previously met the intensity and organization standard under Tadić.¹⁰⁴ Even then, as the FARC–Colombia example shows, sporadic acts of violence may persist after a formal peace agreement. Conversely, as both the FARC–Colombia and US–Al-Qaeda examples illustrate, the near or total absence of hostilities between any belligerent parties may not dispositively terminate a NIAC. Yet, since Article 3 common to the four Geneva Conventions extended IHL's automatic and objective application to NIACs, IHL has not applied to “any form of anarchy, rebellion, or even plain banditry”.¹⁰⁵ The authoritative Commentary to Geneva Convention III (GC III), for example, explained the scope of IHL's application to NIAC as reaching “conflicts [that] are armed conflicts, with ‘armed forces’ on either side engaged in ‘hostilities’ – conflicts, in short, which are in many respects similar to an international war”.¹⁰⁶ Thus, Additional Protocol II specifically excludes from the definition of armed conflict – and the ambit of IHL – “situations of internal disturbances and tensions, such as riots, isolated and sporadic acts of violence and other acts of a similar nature”, which are “not … armed conflicts”.¹⁰⁷ Moreover, international tribunals have repeatedly emphasized the importance of protracted armed violence in assessing the existence of a NIAC,¹⁰⁸ and have similarly insisted on the existence of sufficiently intense armed violence perpetrated by organized armed groups for a NIAC to exist.¹⁰⁹ Because even organized isolated and sporadic acts of violence would not be sufficient to constitute a NIAC, it would be illogical for such disparate acts of violence to ipso facto prolong a NIAC. Likewise, even protracted violence perpetrated by insufficiently organized groups could not ipso facto sustain a NIAC because such disordered violence could not constitute a NIAC in the first place.¹¹⁰ At the very least, the diminution of either organized or intense hostilities below the threshold of the Tadić test is strong – if not conclusive – evidence that a NIAC has terminated.

On the other hand, a legal test for the end of a NIAC that only requires one or both parties to collapse into disorder, or to quit the fight from exhaustion or utter defeat, would tend to prolong situations of armed conflict and inhibit the restoration of peace. Thus, resolving the termination of a NIAC requires evaluation of the diminution of organized hostilities along with both the quality of pacification and the length of time since organized hostilities have diminished below the Tadić threshold. These factors provide a degree of certainty that hostilities have permanently and stably ended.¹¹¹

Element 2: The quality of pacification

The quality of pacification of a NIAC is both an important condition of the termination of a NIAC for legal purposes and an important safeguard against the revolving-door problem. Separate from a diminution of organized hostilities, factors like the degree to which the parties have achieved their lawful objectives, uncontradicted statements of the parties concerning the termination or continued existence of the NIAC, and the implementation of concrete steps toward pacification are important if non-dispositive factors for evaluating genuine pacification and, consequently, the termination of a NIAC.¹¹²

Fulfilment of objectives

The question of whether the objectives sought by the parties upon initiation of the NIAC have been achieved is central to determining whether there is a risk that violence will reignite. Analysis of the fulfilment of objectives is required whether the objectives sought are legal or not; even if those objectives are legal, whether the objective has the domestic and international political support needed to render a permanent change may weigh in the assessment of whether a NIAC has ended. In extreme circumstances,¹¹³ self-determination, for example, is considered a legal objective, but the scepticism expressed by the United States, the United Kingdom, France, the Netherlands and others towards enshrining the principle as a right in Article 1 of the International Covenant on Civil and Political Rights has endured.¹¹⁴ Movements for self-determination that have the backing of powerful States – such as the support for Eritrea from the People’s Republic of China during the Cold War, and from the United States following the end of the Cold War¹¹⁵ – may be more likely to achieve peaceful settlements without risk of resumption. Other movements for self-determination that may well be legally sound, such as those of the Tuareg, the Sikhs or the Baloch,¹¹⁶ have languished. The situation of the Tuareg and other minorities in Mali illustrates that even fulfilling an objective in whole or in part (taking power following abdication of the Malian government in northern cities) cannot result in a permanent change if other conflict parties refuse to accept the change and they have the support of larger powers.

Legitimate fights for self-determination must also be distinguished from campaigns of aggression, terrorism, and narcotics trafficking. The involvement of AQIM with the Tuareg groups’ struggle against oppression – despite different goals – likely doomed international recognition of any of the groups’ objectives.¹¹⁷ Similarly, the FARC's pursuit of self-determination for the indigenous population of Colombia was tainted by its turn toward criminality and its participation in narcotics trafficking.

The clarity of the objective sought is important in assessing pacification. During the peace talks between the Third and Fourth Eelam Wars in Sri Lanka, the LTTE appeared to have dropped its long-standing demand for a separate State and agreed with the Sri Lankan government that a “federal solution” may be possible.¹¹⁸ Although, “[w]ith the benefit of hindsight”, it is clear that “the basic long-term objectives of the conflict parties – secession [versus] maintenance of a unitary state – did not change”,¹¹⁹ the period between February 2002 and April 2003 saw a cessation of hostilities along with peace negotiations,¹²⁰ as a result of which – depending on the temporal factor, discussed below – it may be possible to conclude that the NIAC had ended in February 2002, even if another one began in April 2003.

Statements of parties

The public statements of the conflict parties, although not dispositive without the surrounding elements, are another important indicator of their intention to be bound to their commitments. For example, one of the enduring arguments in support of the position that the NIAC between the United States and Al-Qaeda has ended is that successive US presidents have repeatedly publicly stated over a decade that combat operations in first Iraq and then Afghanistan were over.¹²¹ Despite these statements, however, the continued deployment of US forces and positions of the US government taken in the shadow of litigation have contributed to the United States’ legal assertion that its armed conflict with Al-Qaeda continues.

Implementation of concrete steps toward pacification

Additionally, the implementation of concrete steps towards pacification by the parties is a significant and objectively measurable indicator that a NIAC has terminated without substantial risk of resumption. Such steps may include disarmament, diplomatic negotiations, the issuance of amnesties and the establishment of accountability mechanisms.¹²² These steps entail tangible measures that diminish the likelihood that the parties will resume hostilities. They also tend to promote the establishment of peace by providing a concrete safeguard against the revolving-door problem.

For example, the peace agreement between the FARC and Colombia required the FARC to surrender its weapons and its fighters to demobilize. Importantly, the peace agreement also included a monitoring and verification component, which both corroborated demobilization and disarmament in fact and built confidence between the parties, thereby contributing to the quality of the pacification of the armed conflict.

An instructive counter-example on this factor is the period of time between the First and Second Eelam Wars in Sri Lanka. The First Eelam War “ended” with the 1987 Indo-Sri Lankan Peace Accord, but the exclusion of the LTTE from the negotiations which resulted in that agreement, and the LTTE's subsequent refusal to disarm, deprived the agreement of the concrete steps toward pacification that would mark the end of a NIAC. Thus, the three-year period between 1987 and 1990, when India removed its troops from Sri Lanka, would not constitute a break in the otherwise ongoing NIAC between Sri Lanka and the LTTE. Indeed, that period of quiescence, which might have seemed like peace as a descriptive matter, was in reality a period of continuous LTTE rearmament and reconstitution. Due to the deficient quality of pacification following the Indo-Sri Lankan Peace Accord, therefore, IHL should be considered as continuing to apply to the conflict between the LTTE and the Sri Lankan government as a matter of law between 1987 and 1990.

Element 3: Length of time since cessation – the “five-month timer”

The most difficult but perhaps the most essential element in any test to determine the end of a NIAC is determination of an acceptable period of time to assess the first two elements described above. Under Tadić, the intensity of a NIAC must be “protracted”. While there are differing views on the duration of hostilities sufficient to constitute “protracted” under the Tadić test, International Criminal Court (ICC) jurisprudence, at least, points to a “minimum length” of five months of armed conflict.¹²³ Although the ICC's durational jurisprudence concerning the beginning of a NIAC alone cannot be described as crystallizing customary IHL, we propose adopting an identical period for assessing the end of a NIAC because of its unusual specificity. Such an approach has the benefit of preserving the symmetry embedded in the reverse Tadić theory, which elegantly aligns the onset and termination of NIACs for legal purposes. It also has the benefit of clarity, imposing certainty and facilitating application. The proposed five-month timer would begin upon diminution of organized hostilities below the Tadić threshold, which is contemporaneously measurable. The pacification analysis would then be applied to the five-month period, to conclusively determine whether the NIAC had terminated upon diminution of hostilities below the Tadić threshold.

The importance of including a specific period for any test determining the end of a NIAC cannot be overstated. The inclusion of this factor places an obligation on conflict parties and observers to start a clock on pacification, and it creates an incentive for the parties to maintain peace and keep the clock running. The inclusion of the temporal requirement may therefore contribute to the successful termination of a NIAC in law, which would include resumption of normal governance of the civilian population, which is always the desired outcome, as opposed to decades-long suspension of domestic law in favour of IHL. While the case studies listed above may not have met all elements of this proposed test, it is possible to identify certain windows where an assessment that the NIAC had ended and greater resources towards pacification might have affected the eventual resumption of hostilities – for example, between the Third and Fourth Eelam Wars.

This temporal limit would also help resolve ambiguous terminations of NIACs, especially in complex situations. For example, in the case of the US–Al-Qaeda armed conflict, the prolonged diminution of organized hostilities below the Tadić threshold for NIAC – well beyond the five-month period we suggest – indicates that that armed conflict has concluded. This conclusion is corroborated by the quality of the pacification of the US–Al-Qaeda NIAC. Notwithstanding some contrary statements by US officials, and notwithstanding the absence of a peace agreement between the United States and Al-Qaeda, the United States has achieved its objectives of defeating Al-Qaeda by rendering it, according to the US intelligence community, incapable of attacking the United States. Moreover, the United States has taken concrete steps that suggest the end of that armed conflict by, for example, ending combat operations in Afghanistan, withdrawing all its forces from Afghanistan, and ceasing to carry out drone strikes against core Al-Qaeda forces in Afghanistan or Pakistan.

Other lengths of time may be arguable pursuant to the jurisprudence succeeding Tadić and further analysis of contemporary NIACs, but in any case, we encourage the adoption of the temporal requirement in this proposed test.

Conclusion

As the case studies illustrate, the undeniable rise of non-State actors in power and capability requires a more concrete and methodical approach to the determination of NIAC termination. The codification of a standard to end NIACs is perhaps especially critical when the non-State actors in question are considered by States to employ terrorism, since it is undeniable that State claims of “counterterrorism” to justify suspension of human and civil rights in favour of “looser” NIAC obligations have expanded exponentially over the past two decades. The proposed test would limit the discretion that States currently have in conducting “forever wars” against expanding sets of political opponents. At the same time, constructing a mechanism to restore peace and the primacy of both international human rights law and ordinary domestic criminal law would promote peace and facilitate States’ preservation of the historic distinction between terrorists and legitimate combatants. Most importantly, adopting a concrete standard that establishes an objectively verifiable test for the end of a NIAC may bring relief to beleaguered civilian populations who are the primary victims in such indefinite conflicts. The test proposed here is an initial effort at a practical, fact-based solution for one of IHL's major gaps.