Why do rebel actors increase the intensity of violence at particular points and decrease them at other points? It is hard to exaggerate the policy relevance of this question, which is the central puzzle in Isabelle Duyvesteyn’s new book, Rebels and Conflict Escalation. Its underlying assumption is that if we identify causal factors explaining patterns of conflict escalation and de-escalation in various historical cases of conflict, we can predict the rise and decline in violence in other current and future insurgencies. Duyvesteyn sets out to identify multiple triggers of conflict escalation through careful examination of various historical cases, so that policy makers can keep an eye on them and manage the consequences of their decisions and actions in current and future insurgencies.

A growing literature seeks to understand the shifting dynamics of civil warfare. Competing accounts emphasize “grievance” and “greed” to analyze how dissidents escalate violent action, and although their differences matter for both policy and scholarship, scholars recognize the need to conduct a head-to-head evaluation of competing explanations that would synthesize a systematic body of knowledge on this topic. Rebels and Conflict Escalation attempts such a synthesis via the comprehensive theorization of escalation and de-escalation of rebel violence based on evidence generated from a broad range of case-based evidence, introduced in the book through multiple vignettes. The vignettes offer empirical details of conflict dynamics in cases drawn from Argentina to Sierra Leone, Northern Ireland to the Philippines, from Nicaragua to the Tamil Tigers, and more.

The book, organized thematically into 10 chapters, builds on the Clausewitzian framework of defining war as a politically motivated conflict among people, the state, and rebel forces. The political logic and initial causes of war, however, do not explain escalation, which is theoretically conceived as developing along two main axes: political will and capability. Chapter 2 presents four possible scenarios of escalation based on the interaction of political will and capabilities, which forms the predictive framework of the book, nicely summarized in a 2 x 2 table (p. 204).

The first escalation scenario follows from increased political will coupled with prevalent capabilities, which occurs when there is a change in the political opportunity structure and an increased salience of the issues at stake, as in the case of the Vietnam War. The second escalation scenario follows from an increase in both political will and capabilities, which leads to increased targeting of civilians or indiscriminate killings, generally referred to as “atrocity shift.” The third scenario of escalation follows from pairing an increase in capabilities with unchanged political will. This escalation outcome stems from the substitution or change of tactics, targets, and the geography of conflict. The fourth scenario of unchanged political will and capabilities does not produce escalation and hence is not pursued further. Similar outcome matrices also explain de-escalation of conflict in terms of various combinations of political will and capabilities of rebel groups.

The primary contribution of the book is the formulation of the “causal trajectories” of these four forms of escalation occurring through shifts in extremity, saliency, atrocity, and substitution (see table 10, p. 192). In contrast to previous studies in greed/grievance frameworks, which focused exclusively on atrocity shift as the main predictor of escalation, this book lays out three additional shifts and seven alternate escalatory pathways via actors, issues, instruments, tactics, targets, territory, and the duration of conflict. This thorough specification of escalation outcomes in the four quadrants of a 2 x 2 table is, however, not matched by a clarity of definition and of measurement of the core concept of escalation. Some observable implications of escalation, like civilian targeting or indiscriminate killing, are also listed as causal factors contributing to escalation. If the author did not aspire for a causal account, which requires establishing temporal precedence of causal factors over the outcome, but instead highlighted how escalation/de-escalation co-occur and covary with (not caused by) these factors on an ongoing basis, some of these conundrums could have been avoided.

The critical evaluation presented here should not distract readers from the fact that this book is an immensely valuable contribution in the field, particularly because it dives deep into the empirics of violence during multiple episodes of conflict across the world since 1945 to explain and predict conflict dynamics in a theoretically coherent way. What sets this book apart is that, in its bid to be policy relevant, the author does not sacrifice nuance and complexity while providing parsimonious, elegant, and straightforward policy prescriptions. Without relinquishing the goal of generating theoretically rooted social scientific insights, it valiantly grapples with the intricate, synergistic. and even endogenous processes of interaction among multiple variables. The book is all the more engaging and useful for that. Duyvesteyn also concedes that conflict escalation cannot always be depicted as a rational, linear, and willful choice of rebels; as shown in the survey of historical cases, escalation can be accidental, messy, and even an uncontrollable outcome. This book shows that it is possible to do policy-relevant, theory-building social science research without making unsubstantiated assumptions of the linearity and rationality of conflict dynamics.

The primary weakness of this study follows from a methodological choice made in the book: its cherry-picking of truncated episodes of various cases of conflict as evidence. These vignettes are used to illustrate and even substantiate various pathways of escalation in the study. The crucial question here is whether it is possible that the book, unintentionally perhaps, presents only the evidence that supports the author’s argument while ignoring evidence that would potentially contradict it (confirmation bias). In other words, why do we expect that conclusions derived from these vignettes of certain conflict cases will be generalizable enough to other cases to offer a basis for creating forecasting tools for policy makers? On a related note, because certain historical episodes are used to generate hypotheses about causal processes of escalation, the same cases can hardly be used to test those theories as well. In other words, the theoretical framework in this book needs further testing.

Perhaps in anticipation of this criticism, the author invites peers and future scholars to subject her hypotheses on escalation and de-escalation to further rigorous testing. Fieldwork, specifically interviews with rebel groups, could be a useful tool for exploring the causal processes of escalation. For example, Duyvesteyn theorizes that conflict escalation happens when there are extremity shifts within rebel groups caused by the situational entrapment of rebels. When rebels find no other way out of their lives as outlaws, a hardening of position and escalation ensue. On the flip side, de-escalation follows from rebels’ willingness to put a brake on violence, either due to fear of losing public support and legitimacy or to defection and a loss of foreign sanctuary. Although these hypotheses are intuitively appealing, they can be substantiated through interviews with current and former rebels, which is difficult but not impossible.

This review would be incomplete without probing how the book conceives the role of the state in conflict escalation and de-escalation. Chapter 4 highlights four courses of action open to policy makers responding to violence. The state can use (1) moderate repression combined with concessions, (2) overwhelming and outright force, (3) restraint and nonviolence, or (4) nonresponse (do nothing). However, this chapter offers no insights into why the state responds as it does. Perhaps this is not a question the author is interested in. However, because the book argues that strategic interaction among state and rebel actors shapes conflict dynamics, it is important to theorize drivers of state response. Political imperatives and social fragmentation within states would likely affect state interaction with rebel groups and their propensity to escalate or de-escalate. An apolitical, asocial, and undertheorized state might be an inevitable result of relying on earlier scholarship on interstate conflicts; for example, the idea of threshold as a route to escalation and that of norm convergence as a route to de-escalation are derived from Schelling’s seminal contribution to nuclear strategy in the 1960s. At the very minimum, marrying Schelling’s interstate framework with a Putnamesque two-level approach, which includes a domestic level of analysis, would have been more apt for explaining intrastate violence.

Unless we pry open this black box of what happens inside the state, the dynamics of conflict escalation will remain a mystery. For example, states in conflict zones are known to respond to rebel violence with welfare generosity, which weans supporters from rebel groups. If the state responds to escalation with a favorable response, including expansion of the welfare state, is that a recipe for escalation or de-escalation? In all fairness, the author anticipates some of the critical commentary made in this review about the paucity of empirical evidence and undertheorization of the role of the state and, as mentioned, invites rigorous testing of her hypotheses by peers and future scholars, which will undoubtedly enrich the field and generate new insights for policymakers.

In conclusion, Duyvesteyn chooses a historically attuned approach to predict the rise and fall in rebel violence and approaches the task of generating the policy implications of her research with cautious humility. If some find this expansive and authentic intellectual incursion into complex issues of conflict dynamics imprecise, it is more useful than the alternative of definitive and elegant quick fix policy prescriptions, particularly because it offers avenues for future research. Rebels and Conflict Escalation will be a valuable companion for conflict researchers and policy makers for many years to come.