Routine but Consequential: How Ceasefire Monitors’ Reporting Constructs Opportunities for (Non)Compliance by Conflict Opponents

ABSTRACT

Third party ceasefire monitors routinely gather and report information on conflict events. Although ceasefire monitoring is a common conflict response intervention generally correlated with ceasefire durability, how its routine practices contribute to ceasefire compliance and noncompliance is little understood. This article asks how reporting, monitors’ most common practice, affects conflict opponents. Based on the experiences of more than 100 monitors, as well as archival research, I develop theory for how ceasefire monitors’ reporting constructs opportunities for conflict actors to demonstrate both compliance and noncompliance, and show evidence for this in cases from Kosovo and South Sudan. That monitoring can produce ceasefire noncompliance challenges existing understandings of monitoring as generally contributing to ceasefire durability. The implication is that even credible monitoring and accurate reporting may have inadvertent consequences on conflict trajectories.

KEYWORDS:

• Ceasefire monitoring
• compliance
• reporting
• Kosovo
• South Sudan

1. Introduction

By the time the Organization for Security and Cooperation in Europe Special Monitoring Mission (OSCE SMM) to Ukraine suspended its ceasefire monitoring activities in March 2022, it had published at least 2,887 reports, in addition to its non-public reporting to OSCE member states and conflict parties.¹ Did this reporting matter? For at least one former SMM official, the answer was a resounding yes: ‘Our reports were taken seriously enough that the parties were prepared to argue about [their contents]’.²

This article examines this most prevalent, routinized, and time-consuming practice of ceasefire monitors: reporting. A typical independent ceasefire monitoring mission reports its findings, which are thought to contribute to ceasefire durability by providing credible information, building confidence between conflict opponents, providing a means to de-escalate tensions when incidents occur, mitigating accidents, and creating costs for parties that violate a ceasefire agreement, in both inter-³ and intra-state⁴ conflicts. However, how precisely ceasefire monitoring affects, and potentially influences, conflict parties, and their behaviour, remains only partially understood.⁵ To add to this understanding, this study pays attention to monitors’ practices: what monitors do and say, and how such activities are embedded in a wider conflict context.⁶

In particular, I argue that ceasefire monitoring’s reporting practices are central to any function of monitoring effect or influence. I define reporting as the process by which monitors provide a spoken or written account of what they have witnessed, heard, done, or investigated, and the collation and consolidation of that information in formal and informal reports. In this sense, reporting by ceasefire monitors is a practice: a ‘temporally unfolding and spatially dispersed nexus of doings and sayings’.⁷ Reporting constitutes both doing: the collection of information, and saying: the reports themselves. Although monitoring missions take different forms, reporting activities are common to most such missions, many of which try to standardize and institutionalize reporting practices.⁸ Reporting, I theorize, affects, or aims to affect, conflict adversaries and influences, or aims to influence, their behaviour. Typically, such influence is conceived of as aiding cooperation: reporting is thought to affect ceasefire durability by addressing information asymmetries and mistrust by providing credible information to opponents.⁹ However, how precisely the information function of monitors’ reporting unfolds has received little scholarly attention to date.¹⁰ Often, monitoring is discussed in terms of the costs and benefits for conflict adversaries: such parties accept monitoring because they expect benefits or oppose monitoring because it incurs (or is expected to incur) costs.¹¹ By focusing on reporting and how it might lead to opportunities for conflict parties, I nuance this typical cost–benefit paradigm, and conceive of monitoring as having ongoing and cumulative effects through reporting.

Drawing on insights from interviews with monitors from different missions, archival material, and monitors’ reports, I abductively theorize three means by which the reporting of ceasefire monitoring may affect or influence conflict parties. These means are in part mechanistic; that is, they act (or could act) as a cause in generating particular outcomes.¹²

Specifically, I theorize three means by which reporting either affects and/or provides opportunities for conflict parties. First, monitoring contributes to conflict narratives. Second, monitoring creates opportunities for compliance and noncompliance by conflict opponents. Third, through silences and elisions, reporting reveals monitoring’s limits. I particularly focus on means two as a novel contribution to understanding how ceasefire monitoring produces wider effects beyond any general function of credible information provision.

This article thus makes three principal contributions. First, it aims to build and develop theory for how reporting contributes to conflict opponent compliance and noncompliance with ceasefire commitments. Second, the article foregrounds everyday monitoring practice, which has received little attention in the emerging literature on ceasefires and ceasefire monitoring. Lastly, the article contributes to the ongoing debate about knowledge production in peace and conflict studies, and particularly to the epistemic and representational practices of conflict knowledge, by extending study of an everyday object, the report, in conflict interventions.¹³ This emphasizes the need for further attention to both reports and reporting practices, beyond their descriptive content.¹⁴

This article proceeds as follows.

First, I briefly discuss what we know about ceasefires and ceasefire monitoring, before outlining my abductive approach to research and noting scope conditions for this study.

Second, I argue for why the reporting of ceasefire monitors both relates and is different in key aspects to those of other types of monitors, such as election observers, human rights organizations, or multi-dimensional peacekeeping operations.

Third, I theorize the different means by which reporting has effects and/or creates opportunities for conflict opponents. I then draw on both archival material and insights from interviews with more than 100 ceasefire monitors to demonstrate the presence and operation of these means in two ceasefire monitoring missions, the OSCE Kosovo Verification Mission (KVM), active from 1998–99, and the three linked ceasefire monitoring missions present in South Sudan between 2014 and 2023.¹⁵ These missions are intentionally selected for their historical, temporal, and geographic discontinuities; case selection considerations are discussed in greater detail below. While, in line with abductive thinking, I do not claim that all ceasefire monitoring practice is generalizable from these two missions, the presence of common mechanisms in diverse cases across time and space suggests the presence of a similar logic to monitoring practice, despite important contextual and mission design differences.

Lastly, I outline some of the article’s implications for future research and practice.

2. What We Know About Ceasefires and Ceasefire Monitoring

Roughly a third of the more than 2,000 ceasefires reached worldwide between 1989 and 2018¹⁶ had provisions for third-party ceasefire monitoring: the process of observing, documenting, and reporting on ceasefire (non)compliance by a third-party actor independent of the conflict parties.¹⁷ I define a ceasefire as any arrangement in which one or more conflict parties suspend organized violence along temporal and spatial dimensions.¹⁸ Though ceasefires take many forms, and the parameters of a ceasefire can vary substantially in both dimensions, where third party monitors are present, they usually have a narrow, compliance-focused mandate.¹⁹

In general, credible monitoring improves the durability of ceasefires and reduces the severity of the violations that do occur.²⁰ Statistical evidence positively correlates monitoring as contributing to the sustaining of both inter-state ceasefires²¹ as well as intra-state ceasefires.²²

A range of monitoring mission types exist.²³ Increasingly, missions are being led by regional organizations or non-governmental organizations, as narrowly focused UN ceasefire monitoring missions become less common.²⁴ Where monitoring struggles, a lack of commitment by the conflict parties is often identified as the primary cause. Some studies thus focus on ceasefire breakdown, despite the presence of monitors.²⁵ Others point to the poor design or inadequate implementation of monitoring, in contexts as different as Nepal, South Sudan, Ukraine, and Yemen.²⁶

Across diverse missions, monitoring practices also vary. Monitoring practices can be classified as primarily pertaining to information collection (e.g. patrolling, inspections); analysis (e.g. determining trends); and engagement with conflict and non-conflict parties, including the provision of information to such actors. However, with the recent exception of studies on technological innovation in monitoring,²⁷ few scholars have disaggregated how individual missions’ practices might influence conflict parties and/or ceasefire durability. In particular, the lack of focus on reporting, ceasefire monitors’ most common, and sometimes only, core activity, motivates this study, given reporting’s importance in bringing together these information collection, analysis, and engagement functions.

In keeping with the wider literature on ceasefires, I assume that conflict parties are relatively strategic in their actions.²⁸ While parties often have mixed motives for entering ceasefire agreements and accepting monitoring, I assume that they understand that there are both costs and benefits to such commitments.²⁹ For instance, parties may benefit reputationally from entering into a ceasefire, while accepting monitoring might make defecting from an agreement more costly. However, a conflict party’s calculation of costs and benefits is dynamic, not fixed, and influenced by internal factors, the intent and resolve of opponents, and, as is the focus of this article, by the (in)actions of third parties. In non-definitive ceasefires in particular, uncertainty about the prospects of conflict resumption mean conflict parties are especially keen to preserve a range of options for future action, up to and including returning to full-scale warfare.

3. Ceasefire Monitoring and Other Reporting Regimes

There is a wider literature on the challenges and opportunities of reporting by, among others, multi-dimensional peacekeeping operations,³⁰ human rights monitors,³¹ and treaty compliance regimes.³² Like the reporting produced by these interventions, it may be tempting to dismiss reports of ceasefire monitors and the underlying process that produces those reports as bureaucratic ‘chatter’ of seemingly little importance.³³ Such critiques are similarly sometimes made of the reporting activities of multi-dimensional peacekeeping operations and human rights monitors.³⁴ While there are evident similarities and some common monitoring logics in these other kinds of third-party interventions, two factors set ceasefire monitoring reporting apart: first, the timing and frequency of reporting, and, second, monitors’ discursive options in reporting. These factors are important distinctions in the reporting of ceasefire monitoring as compared to most other reporting regimes.

First, ceasefire monitors are often expected to report quickly on ceasefire violations, given the exigency of circumstances and frequency of incidents. In unsettled conflicts where violations are common, reporting may also occur much more frequently than in other third-party reporting efforts, with monitors concerned to not fall behind on events and maintain the appearance (if not always the reality) of being credibly informed actors. Ceasefire monitors generally reveal their findings to conflict opponents on an ongoing basis, whereas other observers and monitors may have quite different time frames by which to report and disclose information, driven by potentially varying organizational demands, even if they collect data as frequently.³⁵ While in large, hierarchical, organizations like the UN, a single incident might be subsumed in generalized reporting,³⁶ ceasefire monitors often focus on the granular.³⁷ Since, particularly in contexts of preliminary, non-definitive ceasefires, violations remain frequent, reporting on violations can itself become an arena for potential contestation,³⁸ given that it contributes to an iterative series of interactions with conflict adversaries.³⁹

Second, given the likely occurrence of violations, ceasefire monitors tend to focus on matters of noncompliance. Most ceasefire agreements construct negative obligations on conflict adversaries (for example, Conflict Actor A should desist from a certain act), which means that monitors’ reports usually aim to recount the immediate shortcomings of conflict opponents. Consequently, the discursive options open to ceasefire monitors in their reporting are quite constrained: monitors can either identify breaches of agreement obligations and implicitly or explicitly identify or condemn the responsible part(ies), or they can commend parties for their restraint and compliance.⁴⁰ There are fewer comparable opportunities for ceasefire monitors to focus on underlying change or agreed legal standards, as human rights monitors might do in measuring the progress of a state party over time in implementing its international human rights treaty obligations, or as election observers do in assessing the performance of election management bodies with internationally recognized election standards. While there may sometimes be more long-term actions for ceasefire monitors to comment on (e.g. the withdrawal of certain troops or weapons systems from a particular area), the predominant focus of their everyday reporting is individual security incidents.⁴¹ Attention to individual incidents again contributes to the iterative nature of monitors’ relationships with conflict opponents: there are only so many ways in which opponents can be faulted for their noncompliant acts.

Ceasefire monitors, like other third-party actors, often maintain that their role is only to observe and report, and not to directly influence the acts or omissions of conflict parties.⁴² By their very presence, however, ceasefire monitors both explicitly and inadvertently alter the conflict environment into which they are inserted.⁴³ While a comprehensive discussion of the observer effect – that the act of observing or monitoring a phenomenon changes the phenomenon itself⁴⁴ – is beyond the scope of this article, I argue that monitoring is not a neutral act. Normatively, monitors hope to promote compliance, or at least hope to avoid a ceasefire’s total abrogation.⁴⁵ Reporting is one means by which they hope to achieve such aims. Even where reporting might seem cursory or superficial, therefore, there may be deeper significance to the process of reporting itself, and potential for influence on conflict opponents and their future actions. Before discussing the means by which reporting provides such possibilities, I now set out the research design and theoretical approach of this study.

4. Research Design, Theoretical Approach, and Scope Conditions

This study takes a primarily abductive approach, moving back and forth between data and theory.⁴⁶ An abductive approach enables both examining how the data support existing theories as well as how research findings may call for adjusting existing theoretical understandings.⁴⁷ Both interview and archival data underpin this pragmatic abductive theorization, ‘derived from social actors’ language, meanings and accounts in the context of everyday activities’.⁴⁸ While interview data focused on the routine experiences and reflections of monitors, it also helps interrogate the theory underpinning monitoring practice, and how well this aligns with extant theory on credible information provision mechanisms and effects. Meanwhile, archival data helped confirm anecdotal impressions gained in interviews, provide evidence for the specifically theorized mechanisms, and demonstrate how these mechanisms manifest in differing contexts.

I include three scope conditions in this study. The first is to limit analysis to ceasefire monitoring missions that have the capacity to report (i.e. by having the personnel to do so) and which do report (because they are mandated accordingly). The second condition is to limit analysis to missions monitoring non-definitive ceasefires: that is, preliminary ceasefires which are not part of a comprehensive peace agreement that typically goes well beyond security matters.⁴⁹ In contexts of non-definitive ceasefires, the commitment of conflict parties to a conclusive termination of violence is particularly uncertain.⁵⁰

The third scope condition for the study relates to reporting itself: that, whether formally through a joint information sharing mechanism or because of open publication, or informally through routinized contact with conflict parties, the reports or the information they contain must be shared with conflict parties for the means I outline to be operative.⁵¹

5. How Ceasefire Monitors’ Reporting Affects and Provides Opportunities for Conflict Opponents

How might reporting affect and/or provide opportunities to conflict opponents, beyond the classic theorization of promoting cooperation and overcoming information asymmetry? As shown in Figure 1, I develop theory for three such ways in which reporting affects, or seeks to affect, or can prompt, conflict actors.

Figure 1. Three mechanisms by which reporting by ceasefire monitors affects and/or provides opportunities for conflict parties.

Fig consists of text stating reporting by ceasefire monitors...→ contributes to conflict narratives→ creatives opportunities for compliance and noncompliance by conflict parties by establishing expectations; making agreement commitments specific; setting timeframes→ through silences and elisions, reveals monitoring's limits.

First, monitors’ reporting contributes to conflict narratives.⁵² Especially under the pressures of often fast-moving conflict situations, monitors’ daily or weekly reports are not exhaustive.⁵³ Practical constraints, such as the number of reporting officers in a mission, might also limit the information that can be included in any report.⁵⁴ Some missions, like the SMM, establish thresholds by which conflict events are only included if they are corroborated by multiple sources.⁵⁵ Other missions have set different evidentiary thresholds for inclusion of events.⁵⁶ Monitoring missions know that the attention span of international audiences may be limited and often exercise editorial judgments about what can be omitted to keep a report to a desired word or page limit.⁵⁷ Monitors are also sensitive to the demands and expectations of external audiences, such as their institutional headquarters and sponsors. For all these reasons, monitors’ reports are selective. Selections made in reports invariably elevate certain developments over others as significant. By portraying a certain narrative of the conflict when events are often contested and subject to interpretation, monitors effectively choose from among ‘multiple visions’ of what may be happening.⁵⁸ Accordingly, conflict opponents have an interest in influencing such reports so that the narrative that is portrayed is more favourable to them, or, if direct influence over monitoring is unfeasible, then for one or more parties to consider or attempt to offer a counter-narrative to official reports, whether publicly or in private.

Second, monitors’ reporting creates new opportunities for conflict opponents. This can happen in several ways. Reports state or establish expectations of conflict opponents, make agreement commitments more precise and tangible, and set or suggest timeframes by which such actions should be undertaken. This combination of report features constructs both opportunities for compliance and inadvertent opportunities for noncompliance. While expectations might largely be restatements of the underlying ceasefire agreement between adversaries, monitors may add nuance, precision, and emphasis to these expectations, or make the implicit explicit. In so doing, monitors establish subtle, sometimes shifting, hierarchies in ceasefire agreements, by identifying the issues they are most concerned about. Relatedly, some missions, especially in matters considered especially problematic, may make their expectations more tangible by outlining specific actions needed to realize those expectations.

Moreover, monitors’ temporal horizons may vary. While remedial actions by conflict parties are often expected immediately, understandings of imminence and what timetables are feasible may vary. Particularly where implementation requires a series of steps, timeframes may often be extended. In many contexts, original implementation timeframes are often overrun or delayed.⁵⁹ Given the wider difficulties in resuming negotiations, amending overall agreement timeframes may be improbable, leaving it to monitors to indicate the period for action they deem appropriate and/or feasible.

These opportunities for compliance are institutionalized because of the frequency of reporting and the typical discursive strategies of monitors. This can be illustrated by a hypothetical example. Monitors might identify in a report on day one actions they deem need to be taken immediately by the conflict parties. On day two (or whenever the next report is to be issued), they are likely expected, and nearly obliged, to indicate whether such actions have been taken or remain outstanding, if they wish to remain coherent or be seen as credible.⁶⁰

Even if reporting is more infrequent, a similar logic of updating will be at play.⁶¹ Since monitors are aware of their past condemnations or commendations of conflict actors, they must consider, in addition to the circumstances of any new conflict event, whether the tone of a report about that new event should be more robust, or be somewhat moderated in tone. While this could be described as an example of political judgment in monitoring, there is another constraint: the discursive response options available to monitors. While monitors have relatively few active discursive strategies open to them (to condemn or commend conflict actors; silence is also an option, as will be discussed below), as shown in Table 1, conflict parties have a range of possible responses to monitoring criticism.⁶²

Table 1. Possible conflict actor responses to critiques of ceasefire monitors.

Ceasefire monitors critique or condemn conflict actor A	Possible Conflict Actor Responses
	> A accepts responsibility unconditionally
	> A accepts responsibility conditionally
	> A rejects or denies responsibility
	> A blames another conflict opponent
	> A faults monitoring methodology / conclusions
	> A does not respond

Table 1 describes the range of responses open to a given conflict party. (The combination of possible responses increases accordingly in a multi-opponent conflict). The degree to which any response is employed is in part a function of whether a critique by monitors is an individual occurrence or part of a series of ongoing critiques. Parties that have been critiqued or condemned before may wish to avoid being critiqued or condemned again, or too often, and consequently moderate their response to monitors.⁶³ This interaction consequently creates opportunities for compliance that exist explicitly because a third party is present. This is a co-constitutive relationship because, as has already been suggested, the response (or lack thereof) by a conflict opponent can also influence the nature of a future report or response by monitors.

The third way in which reporting provides opportunities to conflict opponents is through its silences and elisions. Reporting reveals what monitoring missions know, but also the limits of their knowledge.⁶⁴ Understanding monitors’ limits is potentially important information for conflict parties, particularly in unsettled conflicts. A silence in a report might imply that monitors are simply unaware of a particular incident, which in a dynamic conflict setting where verification is difficult, could well occur.⁶⁵ Conflict actors could conclude that monitoring missions lack the ability to detect violations because of limited resources, incomplete geographic coverage, or a lack of sophistication in monitoring methodology.⁶⁶ The other possibility is that monitors are aware of a conflict event but decide not to report it. Reporting indicate what monitors are prepared to put on the record as well as the events about which they will be silent.⁶⁷ While monitors may have justifiable reasons for not reporting certain incidents or for eliding particular details, such as a desire to maintain good relations with one or more conflict parties,⁶⁸ or because they are concerned about offending a conflict actor, and/or wish to protect sources, conflict actors can also obtain information from such omissions, including learning what monitors deem too sensitive to comment on or the political constraints under which a monitoring mission is operating. Such silences or elisions are thus meaningful information to conflict actors.⁶⁹

6. Case Studies: Finding the Presence of Mechanisms in Kosovo and South Sudan

Having outlined the means by which reporting creates opportunities for conflict opponents, I now aim to illustrate that these opportunities are more than theoretical possibilities. I draw on reporting experiences from two cases: the KVM and the three linked ceasefire monitoring missions active in South Sudan. I justify the juxtaposition of these cases based on a ‘least similar’ case selection logic.⁷⁰ The KVM and the South Sudan missions, and the contexts in which they were deployed, were obviously different. The conflicts they addressed occurred in different eras and had different trajectories. Monitoring in Kosovo was deployed for less than a year and conducted exclusively by international personnel; in South Sudan, monitors were deployed for more than a decade, and monitored jointly with the conflict parties. The KVM was a huge operation, eventually encompassing over a thousand monitors; the South Sudan mission much smaller. Monitoring structures, personnel, and methods were differently organized and instituted.⁷¹ However, despite these and other differences, the cases are causally comparable. As I will show, common mechanisms can still be discerned, and the logic of monitoring’s effects remains consistent, despite contextual differences.

Before turning to the illustrative examples, I first briefly provide background on the monitoring missions, and how reporting occurred.

6.1. Kosovo

The KVM was established in October 1998, following an agreement between US envoy Richard Holbrooke and Yugoslav President Slobodan Milosevic to deploy a monitoring mission to verify the 23 September 1998 ceasefire, and compliance by Yugoslavia and the Kosovo Liberation Army (KLA) with UN Security Council Resolution 1199.⁷² A further agreement on 16 October 1998 between the Yugoslav government and the OSCE set out the terms of reference, duration, composition, and structure of the KVM as well as the modalities for cooperation between the mission and the Yugoslav authorities.⁷³ The 16 October agreement specified that the ‘Mission will travel throughout Kosovo to verify the maintenance of the cease-fire by all elements. It will investigate reports of cease-fire violations’.⁷⁴ Although the KLA was not a party to the KVM agreement, it effectively accepted the KVM’s mandate and authority to report on violations.⁷⁵

In its operations in Kosovo between November 1998 and March 1999, the KVM issued at least 91 daily reports, further spot reports on developing incidents, and about a dozen weekly or thematic reports spanning a longer period.⁷⁶ None of these reports were ever publicly released; neither the mission’s mandate nor operating procedures explicitly foresaw such publication.

A typical daily report began with a four to six sentence paragraph summarizing the key developments of the last 24 (or 48) hours. Usually, the first sentence of this paragraph assessed the overall atmosphere, by, for example, asserting that the situation remained calm, or that fighting had been ongoing, or that tensions had reduced, or that it was a relatively active day.⁷⁷ Such overall assessments help constitute the first theorized effect: that reports contribute to the construction of conflict narratives.

The fourth section of the daily report addressed the main security incidents that had occurred in the reporting period. It is on this section that this article’s analysis focuses.

6.2. South Sudan

Following the outbreak of civil war in late 2013, ceasefire monitoring in South Sudan began following a cessation of hostilities agreement between the government and the then main opposition group in January 2014. A series of ceasefire agreements between various conflict parties would be reached over the next 5 years. The initial ceasefire monitoring entity, the Monitoring and Verification Mechanism (MVM), was succeeded by the Ceasefire and Transitional Security Arrangements Monitoring Mechanism (CTSAMM) from August 2015 to September 2018, and then the Ceasefire and Transitional Security Arrangements Monitoring and Verification Mechanism (CTSAMVM) from 2018 onwards.⁷⁸ As of 2023, CTSAMVM continues to operate. However, my analysis of reporting focuses on the period up to 2020, as the COVID-19 pandemic significantly altered the mission’s monitoring protocols in subsequent years.⁷⁹

Between April 2014 and August 2018, the MVM and CTSAMM issued at least 118 reports, which, initially, were not publicly released.⁸⁰ In contrast to the KVM, reporting by the MVM, CTSAMM, and CTSAMVM did not use a single format. Developments in the wider peace process influenced the structure and focus of reports, as did changes in mission leadership and personnel.⁸¹ However, despite these circumstances and an evidently different context, as the illustrative examples below set out, the presence of comparable processes can be readily discerned in both cases.

6.3. Illustrating how Monitors Create Conflict Narratives

As several reporting officers noted, there is significant discretion about how a ceasefire violation might be characterized.⁸² A single event could be connected to a wider and perhaps more ominous pattern or be indicative of a wider mood. Or, an event could be adjudged to be an outlier with no broader ramifications or significance. For example, in its report of 2–3 January 1999, the KVM described individual murders as creating wider social unease, without explaining on what basis that view was reached.⁸³ Elsewhere in the report it is stated that there is relative calm and that relations with the Yugoslav army are generally good.⁸⁴ An immediate contrast can be seen in the report of the following day, where calm in Kosovo is now noted in unqualified terms; while incidents causing death are ongoing, wider anxiety is no longer alleged.⁸⁵

Likewise, in South Sudan, in commenting on the prevalence of sexual and gender-based violence (SGBV), monitors asserted that such conduct is ‘apparently tolerated … and demonstrates a complete lack of discipline’.⁸⁶ However, as others have argued, SGBV in South Sudan was a deliberate instrument of war and could consequently be seen as an expression of military discipline rather than indiscipline.⁸⁷ In both cases, deliberate choices by monitors privilege one interpretation of events over others. Over the course of many iterations of reporting, such choices are repeated and thus contribute to broader narratives.

6.4. Evidence for Reporting Creating Opportunities for Conflict Opponent Compliance and Noncompliance

There are numerous examples of how monitoring reports create opportunities for compliance and noncompliance by conflict opponents. Regarding compliance, to evidence the three dimensions: establishing expectations, making agreement commitments specific, and setting timeframes, I identify brief examples from both cases. I then turn to show examples of how opportunities for noncompliance can also be (inadvertently) created.

6.5. Creating Opportunities for Compliance

6.5.1. Establishing Expectations of Conflict Parties

A classic example of establishing expectations comes from 13 January 1999, when the KVM coordinated the release of eight Yugoslav army soldiers held captive by the KLA near Podujevo.⁸⁸ In response to the detention, the Yugoslav army had deployed more soldiers in the vicinity. Following the release of the detainees, the KVM’s daily report stated an expectation that the additional deployed forces return to barracks, adding that the commanding general had been informed of the Mission’s expectation.⁸⁹ Such a demand was not explicitly derived from the underlying ceasefire agreement: instead, it was a new expectation delineated by monitors and formalized by monitors’ reporting.

Another example comes from a series of events at the Serbian Orthodox Devič Monastery in northern Kosovo in 1999, on which at least three KVM reports commented.⁹⁰ In early February, the KVM learned of an apparent robbery at the monastery, for which the Yugoslav police accused the KLA of responsibility.⁹¹ When the police tried to visit the monastery, they were fired upon by KLA forces. The KVM asked the police not to attempt to visit the monastery again until agreement on access had been reached with the KLA. The police accepted this request but warned they would not wait indefinitely. Over the following weeks, the KVM attempted to facilitate an agreement, thus creating new expectations of conflict parties to demonstrate both restraint and cooperation while these ad-hoc negotiations were ongoing. Because monitors chose to focus reporting on Devič, attention to this situation became more than a local concern.⁹²

In South Sudan, the CTSAMVM witnessed the release of 32 child soldiers in Mir Mir, Unity state, in July 2019.⁹³ In its report, CTSAMVM wrote that it ‘strongly encourages others to follow this example’.⁹⁴ In the same report, CTSAMVM also ‘welcomed the order … for forces to assemble in cantonment sites by 31 July, but note that this has not been achieved’.⁹⁵ Clearly implied in these reports are expectations on conflict parties to release other child soldiers, while providing subsequent opportunities to conflict parties to demonstrate compliance, by, in this instance, assembling in cantonment sites.

6.5.2. Making Agreement Commitments Precise and Tangible

Ceasefire monitoring reports can also construct opportunities for compliance when reports make agreement commitments more specific, precise, and tangible. For example, following incidents in which the KVM was refused access to detained persons in January 1999, the KVM asked for unrestricted access to all KLA detention facilities, as well as Yugoslav police and prison custody facilities.⁹⁶ While the underlying agreement with the government of Yugoslavia granted the KVM ‘full freedom of movement and access throughout Kosovo at all times’, it was silent on access to prisons and police cells.⁹⁷ In stating this demand for action in its reporting, the KVM thus made an implied commitment explicit. This discussion in the report opened the possibility for later compliance by a conflict party, when in mid-February 1999, the KVM successfully requested access to a KLA detention facility.⁹⁸

Numerous instances of making agreement commitments more precise can be found in South Sudan’s monitoring reports. In 2014, the MVM identified government forces as not being compliant with provisions of the ceasefire agreement then in effect in relation to the obligation to ‘assist displaced persons and refugees who wish to return to their original areas of abode’.⁹⁹ In 2015, monitors called on the conflict parties ‘to separate and canton’ their forces to contain violence.¹⁰⁰ In 2018, CTSAMM called for the ‘underlying issues in the Nassir area’ of Jonglei state to be addressed, including an obligation to ‘freeze forces in place [in Nassir] and [for the parties] to refrain from any sort of provocative or offensive actions’.¹⁰¹ Also in 2018, CTSAMVM requested that conflict parties provide updates as to the size of their forces, as well as ‘GPS coordinates [and] contact numbers of the commanders on the ground’.¹⁰² Such statements, while grounded in the underlying ceasefire agreements, translated ceasefire agreement principles into tangible, specific actions that the parties in conflict could achieve, or at least understand the pathway by which achievement would be measured.

6.5.3. Setting Timeframes by Which Actions by Conflict Parties Should be Undertaken

In Kosovo, the delayed release of the bodies of civilians killed in Račak in January 1999 from government custody attracted significant attention from the KVM. Although the KVM never set a deadline, it repeatedly pointed to the need for ‘timely’ and ‘quick’ solutions.¹⁰³

Two examples of setting timeframes are presented from South Sudan. In Figure 2, CTSAMVM translated agreement commitments into specific dates for action, against which progress can be measured.

Figure 2. Extract of a timeline from a monitoring report.¹⁰⁶

Fig 2 outlines the actions required by the peace agreement, the dates by which these actions are to be implemented, and an interpreted timetable set out by ceasefire monitors.

A different example of monitors setting timeframes comes from January 2018. In that month, the government’s First Vice President, visited towns in Jonglei state in the northern part of the country. Some clashes between rival forces occurred at around the time the First Vice President was in the town of Yuai. CTSAMM concluded that ‘it would have been helpful to have delayed the visit until after the CTSAMM workshop of 24–26 January or the HLRF meeting of 5 February in order to discuss it with other signatories … and therefore dispel any misunderstanding’.¹⁰⁴

6.6. Creating Opportunities for Noncompliance

As I have argued, reporting can also create opportunities for noncompliance. Consider the reporting of KVM personnel who helped escort an Oxfam vehicle convoy in Bajgora, northern Kosovo in March 1999. In response to the escort, Yugoslav police threatened that they would impede future humanitarian convoys, which the OSCE then reported.¹⁰⁵ Although other acts of interference with humanitarian operations were occasionally reported, by elevating the Bajgora incident to a national report, monitors may have inadvertently opened the door to future noncompliant behaviour by the Yugoslav forces. As a result of the monitors’ interest, the Yugoslav police was now aware that monitors might be more attentive to humanitarian access limitations as compared to their other activities, which might well have had greater military value.

Conversely, non-elevation of monitors’ findings could also incentivize, or at least not discourage, future noncompliant acts by a conflict actor. The KVM reported on apparent military preparations for road and bridge demolitions on highways E65 and 252 in February 1999,¹⁰⁷ but did not express outright concern until a month later.¹⁰⁸ In the meantime, demolition preparations continued, even though it was apparent that such actions would lead to military escalation.¹⁰⁹

Even monitors’ reporting on acts of compliance can generate opportunities for future noncompliance. For example, in early January 1999, the KVM reported that the Yugoslav army cooperated fully with inspections and provided unrestricted access to weapons and equipment at three locations.¹¹⁰ In reporting a conclusion of full cooperation from this subjective determination, the level of intrusiveness monitors needed to make that determination was conveyed to the army, thus providing information about monitors’ methodology and limits.

In South Sudan, monitors reported in April 2019 that ‘they must be careful to differentiate between violence related to the conflict and that which involves cattle and traditional rivalries’.¹¹¹ Without faulting monitors for seeking nuance, such a bifurcation in a complex conflict like South Sudan is clearly problematic and readily open to abuse by conflict opponents seeking to evade accountability.

In October 2022, CTSAMVM monitors were denied access to a cantonment site in Wunaliet.¹¹² While denying access to monitors is not in itself unusual, the reporting process revealed that it had been 20 months since monitors’ last visit to this site. Moreover, a letter of complaint sent following the incident had been ignored. From monitors’ reporting, therefore, other conflict parties thus learned the frequency of monitoring visits to cantonment sites, the likely (minimal) consequences of noncompliance when access was denied, and the inability of monitors to correct such situations. Knowing such information opens the door to future noncompliance by conflict actors, at least in relation to providing access to similar sites.

6.7. Evidence for Silences in Reporting

As noted earlier, silences in reporting can also convey potentially useful information for conflict parties. For example, a mass shooting in Rakovina, southwest Kosovo, occurred on 25 January 1999, with the KVM learning about the incident within days.¹¹³ Although the KVM did report the incident publicly,¹¹⁴ internal reporting did not follow-up, and there was no mention of the event again until well after the mission had withdrawn from Kosovo.¹¹⁵ Although the Yugoslav explanation that the deaths were caused by a traffic accident lacked credibility, the government’s official explanation was never refuted.

Another form of silence can be observed in ceasefire monitoring in South Sudan, where the use of child soldiers is prevalent. Despite this prevalence, the first MVM report on child soldiers came only nine months into the conflict, in September 2014.¹¹⁶ The next report on child soldiers came nearly six months later, in March 2015, despite noting that ‘the recruitment and deployment of child soldiers has been an issue throughout the present conflict’.¹¹⁷ The long gap in attention to the issue might well indicate to the conflict parties that monitors were much less concerned with underage combatants than might be commonly thought, and/or that monitors lacked the capacity to routinely detect the use of child soldiers.

7. Conclusions

The reporting of ceasefire monitoring has often been overlooked in the wider study of conflict and conflict resolution. Both similar, but distinct, from other forms of reporting, there may be more to ceasefire monitoring reporting than first appears. This article has theorized how such reporting both affects and creates opportunities for conflict actors, beyond the purveying of trusted and credible information, and provided brief illustrations to evidence the existence of such effects and opportunities.

7.1. Implications for Research

There are three main implications from the study for ceasefire and conflict research. First, this study shows the value in disaggregating the information function that is often posited as being key to monitoring effectiveness. By taking a step towards elaborating the effects of monitors’ primary, most frequent activity, reporting, we can gain a better understanding of how reporting can have multiple, overlapping consequences, even in the same conflict context, and how parties can react beyond welcoming or opposing monitoring or a ceasefire more generally.

Second, the article shows how reporting links not only to ceasefire compliance but also, crucially, to ceasefire noncompliance, particularly in contexts of non-definitive ceasefires, where the commitment to a ceasefire is especially variable. By identifying common ways by which this can play out in different monitoring cases, the article complements work that focuses on quantitatively led explanations of why ceasefires and ceasefire monitoring succeed or fail, while not necessarily being contingent on the particularities of an individual case and its circumstances. The presence of common processes in diverse cases across time and space suggests there are more underlying logics to monitoring theory then have generally been acknowledged in the ceasefire literature to date.

Third, the article implies that the typical cost–benefit framework used to explain ceasefires could be further nuanced. While conflict parties still incur costs in accepting monitoring, by examining monitoring practice, the focus shifts to conflict party decisions made in an ongoing ceasefire, which reshape the costs, benefits, and opportunities on offer.

Beyond ceasefire research, the article has implications more broadly for work on third party interventions and compliance in a range of settings. While, as has been pointed out, the frequency and discursive options available to ceasefire monitors may differ from those of other monitoring regimes, the article shows how reporting creates multiple and varied opportunities for the subjects of its reports, even when monitoring is conducted credibly. Such processes may also be at work in other forms of third-party monitoring interventions.¹¹⁸ That said, having an opportunity created by monitors’ actions does not necessarily mean an actor will seize that opportunity. The takeaway from this article is that while reporting is often treated as positively associated with compliance, or at worst is thought to be benign and/or ineffectual, the relationship between monitor and monitored is more complex, given the iterative relationship monitoring reporting creates, and the options that are open to conflict parties to respond. Studying relationships through the lens of such repeated interactions could have significant implications for the broader study of cooperation and conflict.

7.2. Implications for Practice

By the very nature of their task, the work of third-party ceasefire monitors in situations of unresolved conflict is delicate. While reporting is an explicitly mandated task, negotiators, conflict parties, and monitors rarely lay out the implications of reporting beyond the issue of attribution for failings. Given that compliance is almost never absolute, and that both conflict actors and monitors have imperfect information, the sharing of monitoring information is valuable beyond its mere descriptive content. While most reporting entities would always claim that their reports are compiled with care and attention, a more explicit recognition of the concurrent or resulting consequences of reporting is not usually contemplated.

Critically, this study suggests that the effects of monitors’ reporting activities may be multidimensional and even contradictory. The same reporting system that can produce credible information about conflict events can also identify opportunities for conflict actors to exploit, or at least consider exploiting. While the idealized role of a monitor as a ‘referee’ that does not affect outcomes remains many practitioners’ normative ideal,¹¹⁹ a more nuanced reading of the agency of monitors themselves is needed. While monitors’ agency can manifest in multiple ways, at least a degree of this agency is constituted by monitoring’s everyday practices, like reporting, concurrent with, and perhaps despite, any other monitoring activity, overarching mandate, or self-conception.

Acknowledgments

The author acknowledges the assistance of the staff of the OSCE Documentation Centre in Prague (DCiP), Czech Republic, at which archival research was conducted in April and May 2023, as well as helpful feedback on earlier versions of this article from Ana Carolina Rodrigues, Johan Karlsson Schaffer, Fredrik Söderbaum, Kilian Spandler, Valerie Sticher, Fisseha Fantahun Tefera, and the anonymous reviewers and editors of this journal.

Disclosure Statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This research was supported by grants from the Adlerbertska Stipendiestiftelsen (grant number AD2022-0172) and the Stiftelsen Karl Staaffs fond (application 3858, 2022). Interviews conducted in 2018–19 were supported by the United States Institute of Peace.

Notes

1 Calculation by the author: 2,432 daily reports, 242 spot reports, 180 status reports, and 33 thematic reports. See OSCE Documents Library, 2014–2022.

2 Interview with former senior SMM official, September 2023.

3 Fortna, Peace Time.

4 Clayton and Sticher, “The Logic of Ceasefires in Civil War.”

5 Pinaud, “Home-Grown Peace,” 471.

6 Bueger and Gadinger, International Practice Theory, 2.

7 Schatzki, Social Practices, 89.

8 Krause and Kamler, “Ceasefires and Civilian Protection Monitoring”; Mackiewicz, “More than Counting Ceasefire Violations”; United Nations, “Guidance,” 53.

9 Fearon, “Rationalist Explanations for War”; Fortna, Peace Time; Grist, “More than Eunuchs”; Joshi et al., “Built-in Safeguards.”

10 Bara et al., “Understanding Ceasefires,” 334; Höglund, “Obstacles,” 214–15; Palik, “Watchdogs of Pause”; Pinaud, “Home-Grown.”

11 Fortna, Peace Time; Pinaud, “Home-Grown”; Potter, “Ceasefire Monitoring and Verification.”

12 Falleti and Lynch, “From Process to Mechanism.”

13 Bliesemann De Guevara and Kostić, “Knowledge Production in/about Conflict and Intervention”; Kosmatopoulos, “Sentinel Matters”; Whalan, How Peace Operations Work; Hellmüller et al., “Knowledge Production on Peace.”

14 Adler and Pouliot, “International Practices.”

15 See Appendix 1 for details of research ethics approval and for a discussion of the research methodology employed in this study.

16 Palik et al., Conflict Trends, 21–2.

17 Bara et al., “Understanding Ceasefires,” 334.

18 Bara et al., “Understanding Ceasefires,” 329; Verjee, “Ceasefire Monitoring under Fire,” 808.

19 Findlay, “The Role of”.

20 Åkebo, The Politics of Ceasefires; Potter, “Ceasefire Monitoring and Verification.”

21 Fortna, Peace Time.

22 Clayton and Sticher, “The Logic of Ceasefires.”

23 Buchanan et al., Ceasefire Monitoring; Martin, “Crossing Boundaries.”

24 Krause and Kamler, “Ceasefires and Civilian Protection Monitoring in Myanmar”.

25 Höglund and Wennerström, “When the Going Gets Tough  …  Monitoring Missions and a Changing Conflict Environment in Sri Lanka”; Loquai, “Kosovo – A Missed Opportunity for a Peaceful Solution to the Conflict?”

26 Palik, “Watchdogs”; Pinaud, “Home-grown Peace”; Verjee, “Ceasefire Monitoring in South Sudan,” “Ceasefire Monitoring Under fire.” Among the issues that affect the implementation of monitoring are the composition of monitoring entities, including who selects monitors.

27 Grand-Clement, “Exploring the Use of Technology for Remote Ceasefire Monitoring and Verification”; Sticher and Verjee, “Do Eyes in the Sky Ensure Peace on the Ground.”

28 Clayton and Sticher, “The Logic of Ceasefires,” 4–5.

29 Clayton et al., “Ceasefire Success,” 349–50.

30 Amicarelli and Di Salvatore, “Introducing the PeaceKeeping Operations Corpus”, Clayton et al., “The Known Knowns and Known Unknowns of Peacekeeping Data”; Duursma and Smidt, “Peacekeepers without Helmets”; Mac Ginty, “Peacekeeping and Data.”

31 Alston, “Hobbling the Monitors”; Bagozzi and Berliner, “The Politics of Scrutiny”; Knoll-Tudor et al., “Cartel of Silence”; Lesch, “Contested Facts.”

32 Charlesworth and Larking, Human Rights and the Universal Periodic Review; Leurdijk, “Fact-Finding.”

33 von Billerbeck, “No Action without Talk?” 478.

34 Charlesworth and Larking, Human Rights; von Billerbeck, “No Action without Talk.”

35 Duursma, “Information Processing Challenges in Peacekeeping Operations”; Martin-Brûlé, “Competing for Trust”; Molony and Macdonald, “Re-evaluating International Observation of Kenya’s 2017 Elections,” 606.

36 Mac Ginty, “Peacekeeping and Data”, 695.

37 Interviews with former SMM officials, July 2023, CTSAMVM official, July 2023.

38 Åkebo, “‘Coexistence Ceasefire’ in Mindanao.”

39 Interview, CTSAMVM official, July 2023; Verjee, “Pleading for Peace”; Young, “Intermediaries.”

40 Verjee, “Pleading for Peace.”

41 Interviews with former SMM monitor, July 2023, CTSAMVM official, July 2023.

42 Interviews with former senior SMM official, July 2022; former senior MVM official, February 2019; Bellamy and Griffin, “OSCE Peacekeeping”; Oberson, “OSCE Special Monitoring in Ukraine”; see also Musu, European Union Policy, 127–128.

43 Åkebo, The Politics of Ceasefires; Verjee, “Monitoring Ceasefires is Getting Harder.”

44 Sassoli de Bianchi, “The Observer Effect”; Spano, “Observer Behavior.”

45 Interview, former SMM Monitor, June 2022.

46 Dubois and Gadde, “Systematic Combining,” 555.

47 Thornberg, “Informed Grounded Theory,” 247.

48 Blaikie, Designing Social Research, 89.

49 Clayton and Sticher, “The Logic of Ceasefires,” 6–7; Clayton et al., “Ceasefire Success,” 345.

50 Wiehler, “Deciding on the Tit for the Tat,” 418.

51 While this study focuses on reports not initially released in the public domain, some monitoring missions produce both public and confidential reporting. Mandates may require monitoring missions to negotiate or agree the content of such publicly available material, thus introducing further opportunities for conflict actors to influence and shape reporting.

52 Sticher and Verjee, “Do Eyes in the Sky.”

53 Interview, CTSAMVM monitoring official, July 2023.

54 Interview, former senior SMM official, September 2023.

55 Interview, former senior SMM official, September 2023.

56 Interview, CTSAMVM official, July 2023; OSCE official, October 2023.

57 Interview, former monitor, July 2023.

58 Potter, “Ceasefire Monitoring and Verification,” 4.

59 Jeong, “Peacebuilding: Conceptual and Policy Issues.”

60 Interview, former SMM Monitors, June 2022.

61 Interview, former CTSAMVM Monitor, July 2023.

62 Interview, former SMM Monitors, May 2023.

63 Interview, former MVM Monitors, February 2019.

64 Interview, former SMM Monitor, July 2023.

65 Interview, former SMM Monitor, September 2023.

66 Pinaud, “Home-Grown Peace”, Potter, “Ceasefire Monitoring and Verification.”

67 Interview, former MVM official, February 2019; former KVM monitor, September 2023.

68 Samset, “Trapped”; Verjee, “Ceasefire Monitoring in South Sudan.”

69 Orlandi, Les Formes du silence; Schröter, Silence and Concealment.

70 Koivu and Hinze, “Cases of Convenience?”, 1024.

71 Maisonneuve, “The OSCE Kosovo Verification Mission”; Verjee, “Ceasefire Monitoring in South Sudan.”

72 Loquai, “Kosovo,” 81; Walker, “OSCE Verification Experiences in Kosovo.”

73 Loquai, “Kosovo,” 81.

74 OSCE, “Agreement on the OSCE Kosovo Verification Mission,” 531.

75 US Department of State, “Ambassador William Walker: On-the-Record Briefing.”

76 Calculations by the author based on review of material held at the OSCE DCiP in May 2023. Reporting was less frequent at the start of the mission, and often, over weekends, only one report was issued to cover both Saturday and Sunday. Reports issued after the KVM was evacuated from Kosovo but before its mandate was terminated in June 1999 have not been included in the count or the subsequent analysis; given the absence of monitors on the ground, these reports largely contain second-hand information.

77 The terms of use of OSCE archival material labelled as "Restricted" are that such material may not be quoted directly. The author has used close synonyms to comply with the conditions of use of OSCE archival material. See OSCE, “Recommendations for Quoting OSCE Documents.”

78 Verjee, Ceasefire Monitoring in South Sudan, 12.

79 Interview, CTSAMVM official, July 2023.

80 Verjee, “To aid South Sudan’s Cease-Fire.”

81 Interviews, former MVM and CTSAMM monitors, February 2019, July 2023.

82 Interviews, former SMM monitors July 2023; CTSAMM and CTSAMVM monitors, February 2019 & July 2023.

83 OSCE, “SEC.FR1/99.”

84 OSCE, “SEC.FR1/99.”

85 OSCE, “SEC.FR4/99.”

86 CTSAMM, “Report 2018/12.”

87 Pinaud, War and Genocide in South Sudan.

88 Greco, The OSCE’s Kosovo Verification Mission, 5.

89 OSCE, “SEC.FR22/99.”

90 OSCE, “SEC.FR79/99”, “SEC.FR80/99”, “SEC.FR94/99.”

91 OSCE, “SEC.FR79/99,” 4.

92 Interview, former KVM support staff, September 2023.

93 CTSAMVM, “Thirteenth Meeting Outcome Report.”

94 CTSAMVM, “Thirteenth Meeting,” 3.

95 CTSAMVM, “Thirteenth Meeting,” 3.

96 OSCE, “SEC.FR67/99.”

97 OSCE, “Agreement on the OSCE Kosovo Verification Mission,” 531.

98 OSCE, “SEC.FR102/99,” 2.

99 IGAD MVM, “V006,” 5.

100 CTSAMM, “Report on Alleged Violations and the Situation in Unity State,” 6.

101 CTSAMM, “Report 2018/13,” 4.

102 CTSAMVM, “Technical Committee Fourth Meeting Outcomes Report,” 6.

103 OSCE, “Racak Victims Remain Unburied;” OSCE, “SEC.FR94/99.”

104 CTSAMM, “Report 2018/08,” 5­6.

105 OSCE, “SEC.FR226/99.”

106 CTSAMVM, “Technical Committee Fourth Meeting Outcomes Report,” 22.

107 OSCE, “SEC.FR129/99”.

108 OSCE, “Kosovo as seen,” 59.

109 OSCE, “SEC.FR129/99”.

110 OSCE, “SEC.FR4/99”.

111 CTSAMVM, “Tenth Meeting,” 3.

112 CTSAMVM, “Report 2022/24.”

113 OSCE, “Kosovo as Seen,” 301.

114 UN Security Council, “OSCE Monthly Report on Kosovo.”

115 OSCE, “Kosovo as seen,” 301.

116 IGAD MVM, “V014.”

117 IGAD MVM, “V031,” 4.

118 Boutruche, “Credible Fact-Finding.”

119 Interview, former senior SMM official, July 2022.