Theorizing conflict opponent noncompliance from ceasefire monitoring

Abstract

This article presents theory for how third-party ceasefire monitoring may contribute to, and even motivate, ceasefire noncompliance. That credible ceasefire monitoring may help produce new violence, even if inadvertently, challenges prevailing positive understandings of third-party observation and monitoring, in which ceasefire monitoring is often seen as a form of intervention that helps conflict parties comply with ceasefire commitments. Four mechanisms by which noncompliance may occur when ceasefire monitoring is present are identified. These are, first, that monitoring provides conflict parties with new signaling options in which violence has a potential communicative function; second, that conflict actors’ noncompliance can seek to influence how monitors operate; third, that conflict actors’ noncompliance can help test monitors’ detection capacities; and, fourth, that noncompliance allows conflict parties to check for audience costs. In outlining these mechanisms, this article suggests that there are multiple means by which ceasefire monitoring, like other interventions, can have unintended consequences.

Introduction

The monitoring of ceasefires is supposed to improve their durability. Scholars argue that third-party ceasefire monitoring promotes cooperation, addresses information asymmetry problems, and safeguards the implementation of ceasefire commitments between parties previously in conflict (Clayton and Sticher, 2021; Fortna, 2003, 2004; Joshi et al., 2017). Monitoring is expected to increase the likelihood of conflict parties being detected and punished for noncompliant acts, thus increasing the anticipated costs of noncompliance, and shifting such parties’ cost–benefit calculations (Börzel et al., 2010; Fearon, 1998; Martin, 1992; Mitchell, 1996). Accordingly, credible, impartial third-party ceasefire monitoring is anticipated to deter ceasefire noncompliance.

In this note, I theorize that even credible ceasefire monitoring can instead motivate ceasefire noncompliance by conflict parties, contrary to conventional rationalist expectations (Fearon, 1998; Walter, 1997, 2009). Despite a growing research program on ceasefires (e.g. Bara et al., 2021; Clayton et al., 2021; Krause and Kamler, 2022; Nathan and Sethi, 2023; Palik, 2021), to date, studies of ceasefire monitoring largely focus on how monitoring reinforces compliance, rather than might lead to noncompliance. While the prevalence of monitoring-motivated ceasefire noncompliance is uncertain, reports from contexts as varied as Sri Lanka and Ukraine, for example, suggest some conflict parties’ noncompliant acts are so motivated (Buchanan et al., 2021: 16; Samset, 2004). I seek to understand a puzzle, denoting here a phenomenon not fully explained by existing theories (Mears, 2017), with both theoretical and empirical implications: why might ceasefire noncompliance occur in the presence of, or be motivated by, even credible third-party monitoring?

Focusing on monitoring missions as a whole as my primary unit of analysis, I theorize that ceasefire monitoring creates exploitable opportunities for conflict parties because monitoring both alters and introduces new relationships among conflict actors. To arrive at this position, I build on two strands of theory: theory concerning the role of third-party monitors and theories of noncompliance.

My contribution is twofold. First, I explain why even credible monitoring may have deleterious impacts on conflict, providing a critical perspective to the existing scholarship about third-party monitoring. My theory explains how noncompliance might occur differently in monitored contexts as compared with contexts without monitoring. Second, I outline four mechanisms through which such monitoring-motivated noncompliance could occur. Monitoring offers conflict parties (1) new, indirect, options for signaling, and opportunities to (2) influence how monitoring operates, (3) test monitors’ detection capacities, and (4) check for audience costs. While efforts to distinguish the theoretical motivations for ceasefire noncompliance are now being made (Sticher, 2022b), the potential role of monitors in such dynamics has so far been undertheorized.

In the next section of this note, I review the theory underpinning third-party ceasefire monitoring. In the following section, I consider the relevance of wider scholarship on noncompliance. I then theorize how the presence and activities of third-party ceasefire monitors change the possibilities open to conflict parties. I conclude with a brief discussion of wider theoretical implications and some directions for further research.

The theory of third-party (ceasefire) monitoring

Like other third-party soft compliance regimes, ceasefire monitors are implicitly expected to affect the context of their deployment (Chounet-Cambas, 2011). Such effects are usually theorized as being constructive contributions to conflict resolution (Buchanan et al., 2021). Credible monitors—those with the resources, personnel, and evenhandedness to accurately monitor ceasefires—are expected to help distrustful actors overcome problems of confidence and create and disseminate credible information (Åkebo, 2013; Clayton and Sticher, 2021). By instituting a monitoring regime that can increase the costs to noncompliance with their ceasefire commitments, rational conflict actors should be deterred from violating a ceasefire (Fortna, 2003; Pinaud, 2021). As in other monitoring contexts, parties are expected to positively modify their behavior when being monitored.

While scholars of other third-party interventions such as election observation have shown the inadvertent effects of monitoring, such as greater incentives for opposition election boycotts (Beaulieu and Hyde, 2009), there is to date little attention on the potentially adverse effects of credible ceasefire monitoring, either theoretically or empirically. Critical scholars have tended to focus on empirical cases where monitoring was clearly flawed in design or execution (e.g. Kolås, 2011; Verjee, 2019). However, since ceasefire monitoring is frequently advocated for as a safeguard to maintain a truce (Clayton and Sticher, 2021; Joshi et al., 2017), yet ceasefire noncompliance is still anticipated despite effective monitoring being in place (Potter, 2004), the puzzle of why ceasefire noncompliance should occur in the presence of and/or in response to credible monitoring is not fully theorized nor answered.

In the next section, I argue that such noncompliance may be theoretically possible and even desirable for conflict parties, even when they have no immediate intention to escalate the conflict nor pursue a decisive military outcome, but because conflict actors may sometimes have objectives other than escalation or decisive victory. To strive for these goals, conflict parties may act to instrumentalize the presence of ceasefire monitors for their own ends.

Understanding motives for noncompliance

A host of motivations affect actors’ noncompliance with rules and regimes, whether at the individual, group, or state level (Chayes and Chayes, 1993; Hafner-Burton et al., 2017; Harsanyi, 1962; Oyanedel et al., 2020; Pruitt, 1964; Simmons, 2000). Whether in relation to treaties or other forms of commitment, the motives for noncompliance are not necessarily the converse of those for compliance (Dai, 2005; Hurd, 1999; Raustiala and Slaughter, 2002; Von Stein, 2005). In the international realm, noncompliance is often theorized to be the result of outright defiance (Terman, 2017; Van den Bossche, 1996), constitute a form of bargaining resistance (Crasnic, 2017), be a pretense at compliance by deception (Bluth, 2004), or be the result of agreement ambiguity (Haas, 2000). Such motives may not be mutually exclusive and may overlap (Carmona et al., 2010).

Understandings of noncompliance as binary (i.e. either absent or present) have largely given way to theories of noncompliance as dynamic, graduated, and occurring along a continuum (Haase, 2019; Jancsics et al., 2022; Overman et al., 2014). Some scholars theorize noncompliance in terms of magnitude, qualifying it as either “procedural” or “substantive” (e.g. Nash et al., 2021). Noncompliance can be justified as “principled,” when, for example, a party acts in response to a legal system that has become deficient or outdated (Haase, 2019), as “involuntary” (Checkel, 2001: 576–577) or “inadvertent” (Chayes et al., 1998: 40), when a party is, for example, inattentive to implementation procedures. Accordingly, motives for noncompliance are imbued with varying degrees of intent and significance.

In the ceasefire literature, Sticher (2022b) recently set out a typology of ceasefire violations centered on conflict actor motivations. She terms these as strategic, retaliatory, spoiling, and localized violations. I generally reframe these as expressions of noncompliance. For Sticher, strategic violations, are those ordered, or at least endorsed, by a conflict party’s leadership when seeking to create a military advantage (see also Sticher and Vuković, 2021). Retaliatory violations are those that respond to an opponent’s violation(s), and which seek to safeguard the continuation of a ceasefire, as acts of reciprocity (see also Wiehler, 2021). Spoiling violations are those committed by factions or lower ranks of a conflict actor seeking to undermine or challenge the efforts of their leaders. Localized violations are violations unconnected with conflict party leaderships’ strategic actions, but which occur when combatants on the ground have their own interests, often driven by local circumstances, for engaging in hostilities. Sticher (2022a: 159) also points out that some ceasefire violations are intended to foster compliance, notably when retaliatory acts aim to restore a conflict equilibrium.

Accepting that in the case of retaliation a violation may not necessarily be motivated by noncompliant intent, one notable absence from Sticher’s typology are noncompliant acts that occur beyond a conflict actor dyad, such as in relation to or resulting from the presence of ceasefire monitors.

Here, clear theoretical parallels are applicable from other scholarly literature. For example, some business scholars argue that a third-party regulator’s assessment of a company’s degree of noncompliance could provide insights to competitors who seek to evade those forms of regulation (Bird and Park, 2017; Parker and Nielsen, 2011). Similar dynamics may exist in conflict contexts, where a third-party monitor’s assessment of a conflict party’s noncompliance is potentially valuable information to other conflict parties.

Accordingly, monitoring both alters and introduces new relationships among social actors. In altering or constructing these relationships, four mechanisms by which noncompliance could occur because of the presence and operation of monitors can be theorized. Following Illari and Williamson (2012), I understand a mechanism for a phenomenon (in this case, ceasefire noncompliance) as consisting “of entities and activities organized in such a way [as to be] responsible for the phenomenon” (p. 120). In this approach, identifying mechanisms provides a “how-possible explanation,” outlining how the phenomenon could in principle be produced (Hedstrom and Ylikoski, 2010: 52).

First, credible monitoring provides new communicative, signaling options to those in the conflict setting. Monitors become a communicative go-between, changing the relationship (in an ideal type, two-party conflict) from a dyad to a triad (Simmel, 1950). However, enhanced communication may not necessarily lead to greater compliance, nor necessarily overcome mistrust between opponents. Parties may enact noncompliant behavior, instances of which they anticipate will be, or that they ensure are, observed by monitors, to signal intent and resolve to other parties.

Second, once monitoring missions are introduced, they become actors whom other parties may seek to influence, collectively or individually (Siltaloppi and Vargo, 2017). Whatever the underlying motivations or broader strategy of conflict actors, acts of noncompliance are a way for such actors to influence third parties, who are predisposed to report on incidents, rather than on silences (Schröter, 2013).

Third, parties may seek to test monitors’ detection capacities to determine the precision, reach, and timeliness of monitoring, aiming to gain valuable information to inform their future actions.

Fourth, as actors may be unconvinced about maintaining ceasefire commitments in the long term, they may wish to keep their options open, including to escalate violence or even to abrogate a ceasefire agreement. Concurrently, actors may recognize that there are potential costs to escalation and/or abrogation. They may use monitors to test for audience costs to noncompliance (Richmond, 1998), to determine whether external resolve to impose costs exists, is largely rhetorical, or is mostly absent.

Theorizing how ceasefire monitoring motivates ceasefire noncompliance

Having sketched out these mechanisms, I now theorize them in specific relation to ceasefire monitoring. The presence of these mechanisms does not necessarily mean that conflict actors want conflict to resume in the short-term, nor that they are uncommitted to a ceasefire generally. Rather, ceasefire monitoring should be understood as introducing a fundamental change to ceasefire dynamics that can promote or provide opportunities for ceasefire noncompliance as well as compliance. This is, to date, rarely acknowledged in the conflict studies literature.

I do not claim that these mechanisms are present or operative in all monitored ceasefires, are necessarily likely in all circumstances, nor that they must function entirely independent of each other. Rather, I suggest that they provide plausible accounts of why noncompliance could occur as a direct result of monitors entering and changing the conflict environment. I now elaborate the mechanisms’ theoretical operation in greater depth.

Mechanism 1: providing conflict actors new signaling options

In this first mechanism, effective monitoring gives conflict parties new, indirect, communicative options. While improving communication between parties in dispute is usually thought to reduce uncertainty and promote cooperation and compliance (Fortna, 2004; Haftel, 2007; Keohane, 1984; Morrow, 1994), certainty can be double-edged. Given the dynamism of many conflicts, the mere fact that information is credible may not necessarily lessen a conflict party’s distrust of its opponents, especially if more certain information suggests the opponent is indeed untrustworthy.

In this reading, ceasefire monitors can inadvertently be instrumentalized by conflict parties precisely because monitors can create and transmit credible information. The presence of monitors increases conflict parties’ communicative and signaling options compared with a situation where monitors are absent, as another agent (the monitors) enters the web of conflict party relationships.

How might communicative or signaling options shift because of the presence of monitors? Clearly, a direct attack by a conflict party on an opponent likely remains very costly as it risks retaliation and perhaps undesired escalation. However, a conflict party’s attack on monitors or monitoring infrastructure could still signal to the opponent a conflict party’s capabilities, and obviously be of concern to monitors themselves, and thus likely to be reported by them. Alternatively, conducting activities that conflict parties expect monitors will observe and report, even if they are not a victim of, can also provide a signal of intent and resolve, or of strategic ambiguity, to a conflict opponent. Violence can thus be a communicative action, the transmission of information about which monitors can facilitate. Accordingly, the credible information function expected of monitors can be used by conflict parties for tactical or strategic reasons other than conflict diminution. More interaction is not necessarily constructive interaction (Scott, 1977). The presence of monitors thus offers an additional, indirect channel of communication to conflict parties, the use of which may bear a cost different to that which may be incurred in a direct engagement between conflict parties. Third-party monitoring opens a potentially useful signaling opportunity for conflict parties, even for those that are generally intent on conflict diminution or termination.

Mechanism 2: seeking to influence how monitoring operates

The second mechanism through which conflict actors may engage in ceasefire noncompliance given the presence of third-party monitoring concerns aiming to influence monitors during ceasefire implementation. There are several reasons why conflict parties might seek to influence monitors; such attempts at influence can be aimed at different levels of a monitoring mission. Influence might be deployed to shape the character of a monitoring mission as a whole, by steering monitors toward or away from certain provisions of a ceasefire agreement. If, as in Ukraine, for instance, one conflict party hoped monitors would be less preoccupied with the monitoring of heavy or aerial weapons systems, then occupying monitors with ceasefire violations using other types of arms might be diversionary, or vice versa (Sticher and Verjee, 2023). Influence might also be about pushing individual monitors or groups of monitors to operate in certain ways, such as limiting or increasing the hours during which monitoring is undertaken or pulling monitors toward (or away from) deployment in a geography of concern to a conflict party.

A desire for influence might also arise because of the institutional architecture of the monitoring mechanism; for example, where monitors report to an ongoing mediation process, a demonstration of noncompliance may help a conflict party assert political or military relevance (Höglund, 2011). As it is ostensibly their existential justification, third-party monitors are especially sensitive to ceasefire noncompliance. Consequently, undertaking noncompliant activities would be a logical resort for conflict parties who seek to gain the attention of, and possible influence over, these monitors.

Mechanism 3: testing monitors’ detection capacities

The third mechanism by which conflict parties may engage in ceasefire noncompliance resulting from the presence and activity of monitors is motivated by conflict parties’ desires to determining monitors’ capacities. Particularly in initial phases of the introduction of ceasefire monitoring, as well as in contexts of protracted conflict where monitors may be present for an extended period, I theorize that conflict parties may be uncertain about the capacities and sophistication of ceasefire monitors, which may also change over time. Conflict parties may also wish to understand the timeliness of monitoring and reporting (Pinaud, 2021; Sticher and Verjee, 2023). In terms of deciding whether to risk being held responsible for a more serious offense, it is useful information for conflict parties to know whether monitoring capacities are equitably or unevenly distributed across the conflict theater, and if so, how (Grist, 2001), as well as how sophisticated such capacities might be. A similar mechanism is at work for parties to other areas of regulatory compliance. For example, some states have tested the International Atomic Energy Agency’s capabilities to draw conclusions about the state of their nuclear weapons programs (Furhmann and Berejikian, 2012; Triggs, 1976).

Conflict parties may thus act in a noncompliant manner principally to test the capacities of monitors, and not because of immediate adversarial considerations toward opponents. Knowing whether monitors can monitor some classes of behavior more effectively than others is useful for several purposes. Conflict parties may attribute value to knowing whether non-compliance of a certain type (i.e. the use of a particular weapons system) or in a certain geography will be observed, and to what level of detail. While some may use such information to consider cheating on the ceasefire agreement, there could also be less nefarious purposes for such behavior such as a conflict party seeking to determine the likelihood of a similar act by another conflict party being detected.

Internally, conflict parties, especially those that may lack internal cohesion, may wish to demonstrate to their affiliates and subordinates that monitoring is sufficiently robust to capture violations, and thus deter further spoiling violations of the type Sticher describes. This mechanism in which conflict parties seek to test monitors’ detection capacities may be more prevalent in environments where innovative and more intrusive technology is introduced by monitoring missions, especially if parties are, at least initially, unfamiliar with such monitoring tools (Hug and Mason, 2022; Sticher and Verjee, 2023). Another reason for the existence of such a mechanism is that conflict parties may understand that monitors are constrained logistically or by limited contextual understanding: to what degree these constraints exist and are manifest can only be tested if a ceasefire violation occurs and is reported (or not).

Mechanism 4: testing for audience costs to acts of noncompliance

This fourth mechanism concerns another form of test by conflict actors: to see whether there are actual costs to their ceasefire noncompliance, and if so, what these costs are and of what severity. While in most ceasefire contexts, monitors are not expected to directly impose audience costs (Buchanan et al., 2021), monitors may be empowered or tacitly expected to recommend measures to be taken by other third-party audiences in response to ceasefire noncompliance (Dai, 2002). It is usually up to such third-party audiences to determine whether costs are imposed. However, the information created and transmitted by monitors to these audiences, as well as the interpretation that monitors may give to such information, may be consequential in determining whether the imposition of costs is considered, if not effected (Nutt and Pauly, 2021; Stedman and Rothchild, 1997).

Since conflict parties are assumed to generally want to avoid audience costs but may be uncertain about whether they face genuine peril following their noncompliance, one way to check is what amounts to real world experimentation by trial and error. Conflict actors may thus test monitors to see what they might do when faced with lesser noncompliance, what monitors might do with such information, and under what circumstances they consider advising other actors to escalate their response and impose costs, and to what extent.

Certainly, this mechanism may only partially operate in most circumstances, and monitors may have limited influence over those with the agency to impose costs. A desire to avoid being seen as punitive, or a lack of consensus among international actors, might impede or make hesitant those in a position to impose significant costs to do so. But to see where the red lines truly are, a conflict actor may need to get closer to the line. Knowing whether costs are likely, or can be discounted, would itself constitute valuable information, especially if other factors should eventually make a ceasefire’s continuation more tenuous.

Conclusion

I conclude with two broader points on the wider theoretical implications of this note, before suggesting some directions for further research.

First, although third parties like regulators, overseers, or watchdogs exist in a range of international arenas, from treaty compliance to trade disputes to human rights, theorizations of how third parties might themselves motivate noncompliance with a given regulatory regime are largely absent from debates over why actors comply or do not comply. Although the mechanisms by which such noncompliance might be generated in other settings may differ from those theorized in this article, this does not mean that such mechanisms are theoretically implausible, nor indeed absent, nor dissimilar to those outlined here. Theorizing and testing for their existence could help improve expectations of risks and party behavior more broadly across international affairs.

Second, the application of theoretically grounded noncompliance scholarship more holistically to conflict studies might help to better explain and understand the practices of both conflict parties and conflict intervenors. Noncompliance with commitments should be understood as an expected, rather than only deviant, behavior. Consequently, a broader and more critical application of noncompliance thinking, beyond intransigence or defiance, could broaden the theoretical possibilities for conceptualizing why conflicts persist and recur. Such conceptualizations could benefit from the insight that the presence of mechanisms that generate noncompliance does not necessarily prevent the operation of mechanisms that promote compliance.

For conflict scholars, conceiving of noncompliance as a feature of third-party ceasefire monitoring also establishes a substantively larger research agenda. Establishing the prevalence and magnitude of monitoring-motivated or monitoring-derived noncompliance is one starting point. While there are different types of ceasefire, I expect the mechanisms outlined here to be potentially operative in all kinds of ceasefires in which monitors are present, although the prevalence of a particular mechanism may vary. A deeper interrogation of conflict actors’ motivations, justifications, and actions would help test for the presence and operation of the mechanisms theorized in this article across a variety of historical and contemporary cases, potentially re-historicizing explanations for ceasefire failure and conflict recurrence. For qualitative, case-study led scholarship more broadly, such theory expands the range of less visible possibilities for how third-party interventions affect conflicts, to the point of incentivizing new violence, even when such interventions are not obviously deficient, and such violence is instrumental rather than escalatory. The theory also draws attention to studying the informational effects of third parties in conflict settings. In quantitative approaches, conceiving of noncompliance as an intervening variable, rather than as only an output, in studies that consider why ceasefires endure and fail could nuance thinking beyond the general assumption that the presence of monitoring is always a constraint on future violence. Greater attention to these unanticipated or underrecognized implications of monitoring interventions could help advance more fine-grained thinking about the causes and consequences of conflict.