Roundtable 5-1 on Warlords: Strong-Arm Brokers in Weak States

Warlords coverHow can we understand the important phenomenon of modern-day warlords, often associated with state failure and transborder criminality even as state leaders frequently rely upon them as a source of order or peace in the most difficult of conditions? Kimberly Marten’s Warlords: Strong-Arm Brokers in Weak States blazes a new trail in answering this question, adopting an explicitly inductive approach to theory-building through the study of four important cases of warlordism: Pakistan’s Federally Administered Tribal Areas (FATA), Georgia’s strongmen in Ajara and the Kodori Gorge, the Sons of Iraq, and Russian President Vladimir Putin’s creation in Chechnya, Ramzan Kadyrov.

H-Diplo | ISSF Roundtable, Volume V, No. 1 (2013)
Kimberly Marten. Warlords: Strong-Arm Brokers in Weak States. Cornell University Press, 2012. ISBN: 978-0-8014-5076-1 (hardcover, $35.00).
H-Diplo/ISSF Editors: Thomas Maddux and Diane Labrosse
H-Diplo/ISSF Web and Production Editor: George Fujii
Commissioned for H-Diplo by Thomas Maddux
Published by H-Diplo/ISSF on 4 March 2013
Introduction by Henry E. Hale, The George Washington University. 2
Review by David M. Edelstein, Georgetown University. 5
Review by Matthew Evangelista, Cornell University. 8
Review by William Reno, Northwestern University. 12
Author’s Response by Kimberly Marten, Barnard College and the Harriman Institute, Columbia University  17

Introduction by Henry E. Hale, The George Washington University

How can we understand the important phenomenon of modern-day warlords, often associated with state failure and transborder criminality even as state leaders frequently rely upon them as a source of order or peace in the most difficult of conditions? Kimberly Marten’s Warlords: Strong-Arm Brokers in Weak States blazes a new trail in answering this question, adopting an explicitly inductive approach to theory-building through the study of four important cases of warlordism: Pakistan’s Federally Administered Tribal Areas (FATA), Georgia’s strongmen in Ajara and the Kodori Gorge, the Sons of Iraq, and Russian President Vladimir Putin’s creation in Chechnya, Ramzan Kadyrov.

This engagingly written book makes a number of major arguments, perhaps the most central of which being that warlords are inherently anti-state. While they are capable of providing some public goods, they are necessarily predatory and hardly ever turn into state-builders like Charles Tilly’s wielders of local violence in centuries past.[1] This is fundamentally because world territory is already now completely carved up into states that are claimed and recognized by the international community. In this context, warlords are largely dependent on some kind of state actor for resources and opportunities even as they balk at full loyalty to any state, always seeking to keep their options open and working with various foreign and local forces to maintain their own power and sustain the disorder that makes them valuable. This dependence, however, ultimately makes warlords more vulnerable even to weak states than is commonly thought, leading Marten to suggest that policymakers are often making the wrong decision when they decide they need to accommodate warlords rather than challenge them and undertake serious state-building.

The reviews presented here are unanimous in calling these arguments pioneering in the study of warlordism, likely framing a debate for years to come on a subject about which there is as yet relatively little theory. They also tend to agree that Marten gives good cause for policymakers to at a minimum think very hard before resorting to reliance on warlords, including in places like Afghanistan. Her case studies also come in for praise, representing stimulating and revealing process-tracing accounts of the emergence of different sets of warlords (and the downfall of some) and generating important insights about what William Reno calls their “essential relationship” to states.

While generally very positive in their assessments, the reviewers do raise a number of questions and the occasional direct challenge to   different aspects of this argument. Not everyone is convinced, in particular, that states really have much choice in whether to engage warlords. Both Reno and David Edelstein note that Marten’s lone case of a state successfully ousting warlords in a state-building exercise is Georgia. Edelstein argues that this success might not owe mainly, as Marten has it, to the fact that President Mikheil Saakashvili had the will and a good strategy for removing the Ajar and Kodori warlords after the Rose Revolution whereas his predecessor Eduard Shevardnadze did not, but instead to the possibility that the Georgian state had been growing in strength and that Saakashvili merely culminated a state-building process that would have taken its course anyway. Reno points to negative cases like Mali, where state-building efforts collapsed dramatically and undid years of painstaking work in a matter of days, raising the question of how replicable the Georgian experience is. Both Edelstein and Reno suggest that Marten might have addressed these questions through inclusion of certain kinds of control cases in her research design. Matthew Evangelista also questions another conclusion that Marten draws from the Georgian case, that the opaque and somewhat unsavory methods used by Saakashvili to oust his country’s warlords may not have been possible in   a liberal democracy. Evangelista is more optimistic about democracies’ ability to replicate such feats, noting that they have often proven more than capable of engaging in morally questionable activities as well as spending large amounts of public treasure for the sake of national unity.

Both Reno and Edelstein also find missing in Marten’s account serious consideration of whether warlords might in fact wield substantial popular support or legitimacy and whether this might matter. Reno ventures that one reason so many states have historically accommodated warlords even when they would seem to have had the material resources and will to have ousted them in a state-building project is that “a lot of warlords were popular and a lot of state-building projects were not.” Edelstein would have liked to have seen in the book a rigorous discussion of legitimacy, which he argues warlords can have despite what he interprets as Marten’s view that they are illegitimate “almost by definition.” He also speculates that states might gain or lose legitimacy depending on how they handle warlords.

While Edelstein and Reno appear more reluctant than Evangelista to accept Marten’s argument that even weak states have more power to escape warlordism than is commonly thought, both credit Marten for framing the question and firing an opening salvo in a way that will spur further research capable of answering such questions more definitively in the future



Kimberly Marten is a Professor of Political Science at Barnard College, Columbia University, and Acting Director of Columbia’s Harriman Institute for 2012/13. She is a life member of the Council on Foreign Relations.   Warlords: Strong-Arm Brokers in Weak States (Cornell, 2012), is her fourth book. Her first book, Engaging the Enemy: Organization Theory and Soviet Military Innovation (Princeton, 1993), received the Marshall Shulman Prize. She has also written in Armed Forces and Society, International Security, the Journal of Intervention and State-building, Post-Soviet Affairs, and PRISM (National Defense University), among other sources, and has completed two open-source contract projects for the Pentagon’s Director of Net Assessment. Her current research examines the integration of patronage-based militias into state security forces, and asks under what conditions warlords decide to accept legal governance structures.

Henry E. Hale (Ph.D. Harvard 1998) is Associate Professor of Political Science and International Affairs at George Washington University. He is the author of the books Why Not Parties in Russia? Democracy, Federalism and the State (Cambridge, 2006), a winner of the American Political Science Association’s (APSA) Leon D. Epstein Outstanding Book Award for 2006-07, and The Foundations of Ethnic Politics: Separatism of States and Nations in Eurasia and Beyond (Cambridge, 2008). He is now completing, among other projects, a new book tentatively titled Great Expectations: Regime Dynamics and Patronal Politics in Eurasia.

David M. Edelstein is Associate Professor of International Affairs and Government in the Edmund A. Walsh School of Foreign Service and Department of Government at Georgetown University.   He also serves as chair of the School of Foreign Service faculty and a core faculty member in Georgetown’s Security Studies Program. He received his Ph.D. from the University of Chicago and has received fellowships from Stanford University, Harvard University, and the Woodrow Wilson International Center for Scholars. He is the author of Occupational Hazards: Success and Failure in Military Occupation (Cornell University Press, 2008). He is currently working on two projects: one on exit strategies from military intervention and another on the time horizons of political leaders in international affairs.

Matthew Evangelista is President White Professor of History and Political Science in the Department of Government at Cornell University.  His most recent book is Gender, Nationalism, and War: Conflict on the Movie Screen (Cambridge University Press, 2011).  His current project is a study of the Allied bombing campaign in Italy during World War II, focusing on the attitudes of political and military leaders, bombers and pilots, and Italian civilians towards “collateral damage.”

William Reno is a Professor in the Department of Political Science at Northwestern University. His major publications include Corruption and State Politics in Sierra Leone (Cambridge, 1995), Warlord Politics and African States (Lynne Rienner, 1999), and Warfare in Independent Africa (Cambridge, 2011). His current research investigates causes for variable degrees of organizational cohesion among armed groups in socially divided societies and is based upon field research in Somalia and South Sudan.

Review by David M. Edelstein, Georgetown University

Kimberly Marten has written an important book that is likely to become the first stop for future researchers of warlordism. She investigates the causes and consequences of warlordism and assesses her arguments with a variety of case studies drawn from South-Central Asia and the Caucuses.   She argues that warlords succeed with the support of both central authorities in the state that they inhabit and external patrons, but such support often undermines long-term goals of peace and security. Marten describes the book as a theory-building exercise, and the final chapter provides a variety of hypotheses about warlords from which future researchers may benefit.

In all, it is a clear, engaging, and well-researched book that addresses questions that are central to both contemporary policy debates and larger theoretical questions in the study of political science. In terms of policy, one need only look to Afghanistan or Somalia to understand the importance of the question of warlords. As the world continues to worry about the emergence of other dangerous, ungoverned spaces, the prospect for future engagement with warlords appears high. From a theoretical perspective, political scientists have become increasingly interested in alternative forms of sovereignty that depart from the standard legal notion of state sovereignty that has long dominated the study of international relations. Marten’s book speaks directly to both scholars and policy makers, conveying both lessons and intriguing observations about the nature of sovereignty in some of the most challenging parts of the world.

In the course of this review, I hope to make five major points about the argument that Marten presents in the book. Some of these may be seen as critiques; others are meant to engage Marten’s arguments and push the implications of those arguments. As with so many good books, Marten’s leaves the reader at the end with many more questions than it may have answered itself.

First, Marten correctly observes that, “Stability is not synonymous with security,” (61) but missing from this discussion is a third concept: legitimacy. Leaders may engage warlords in the interest of short-term stability, but such stability does not necessarily carry with it long-term security.   In fact, Marten suggests good reasons to fear that the short-term “convenience” (200) of warlords may fundamentally undermine long-term efforts to build peace and security. Curiously missing from her analysis, however, is any explicit discussion of the legitimacy of warlords or the states that rely upon them. It seems that the only way that stability could become security is if warlords could build legitimacy among the populations they rule over. Marten seems to assume that warlords lack legitimacy almost by definition, but what prevents warlords from working to build the type of legitimacy that could enhance both security and stability? Alternatively, Marten addresses the dangers that a state accepts by doing business with warlords, but she may not go far enough in examining the corrosive effects on a state’s own legitimacy that farming out sovereignty over part of a country may have.

The tradeoff between stability, security, and legitimacy also suggests a fundamental question about the time horizons of the political actors involved in these situations.   Stability is often a short-term concern while security and especially legitimacy would appear to be long-term concerns. If the decision to engage warlords involves providing stability at the possible increased expense of security later on, then the real challenge here becomes how to extend the temporal horizons of leaders engaged in these situations. If this is not possible, —and I expect it would be quite challenging—then it does not augur well for long-term security in states in which warlords have achieved considerable power and authority.

Second, Marten insists on the idea that states have a choice as to whether or not to engage with warlords, but her argument here is less than compelling. This choice is almost certainly highly circumscribed by the capabilities that the state has to govern a certain piece of territory.   The one example she includes that is intended to demonstrate the range of choice that leaders have is the case of post-cold war Georgia in which one Georgian leader, Eduard Shevardnadze, chose to accommodate a warlord while his successor, Mikheil Saakashvili, pursued an alternative strategy of resisting the warlords. The case study is well done and compelling, but missing is sufficient context about the capability of the state to resist warlordism within Georgia.   Saakashvili presumably inherited a more mature state than Shevardnadze did, and perhaps this difference in the development of the Georgian state contributed to a natural evolution away from a reliance on warlords. In other words, perhaps the shift away from warlords in Georgia had less to do with the individual proclivities of certain leaders than it did with other changes occurring in Georgia or the region as a whole. What would have bolstered Marten’s analysis here is another case in which a state had this supposed choice of whether or not to rely on warlords and consciously chose not to do so.   All of the other cases that Marten looks at are those cases in which warlords were either previously or currently engaged, and additional negative cases would have added to the strength of the argument that relying on warlords is a choice, not a necessity. As it is, this claim is a bit unsatisfying, leaving the question of the consequences of warlordism better answered than the reasons as to why leaders rely on them.

Third, and related to my previous points, while Marten’s research design is clear and the case studies are well done, the case studies sometimes lack sufficient context necessary to understand the decisions made to rely on warlords. Most importantly, some of the cases discussed involve situations in which an established state such as Georgia chooses to engage with warlords. Another of the cases, Iraq, takes place in the context of a foreign military occupation by the United States in which the real interlocutor with the warlords is the intervening power rather than the sovereign government. The expediency of relying on warlords may be more attractive to an occupying power seeking an expeditious exit, while the leader of an established sovereign government may be more reluctant to continue to engage with warlords who fundamentally undermine the legitimacy of the central government. The theory-building project in which Marten is engaged might have benefited from a discussion of the effects of various contextual factors that affect the role of warlords. If one conceives of the role of warlords as essentially a bargaining game between a central government and local warlords, then these contextual factors are likely to affect the bargaining leverage of the different sides.

Fourth, Marten’s definition of warlord is sensible enough, but the cases leave some question about the line between warlord and tribal leader. Marten defines warlords as “individuals who control small pieces of territory using a combination of force and patronage.” (3) They “rule in defiance of genuine state sovereignty but through the complicity of state leaders.” (3) As a definition of an abstract concept, this accords with intuitive understanding of what a warlord is. The case studies in the book, however, only serve to raise questions about this definition, especially the meaning of the word “control.” Perhaps most puzzling is the case of the ‘Sons of Iraq’ program which brought tribal leaders together in Iraq in the interest of security as detailed in the lengthy final case study in the book.   This reader leaves that case study confused about the distinction between tribal leaders and warlords.   Engaging local leaders in the messy day-to-day of politics, especially in a society divided by conflict, is surely part of the essential process of moving such a state and society forward, yet does all such engagement qualify as engaging with warlords? Do all local leaders who emerge out of war and who may question the authority of the central government qualify as warlords?   If they do, then does the concept not lose some of its analytical utility? If not, then where does one draw the line between warlord and local, tribal leader?

Fifth and finally, the book delivers a sobering message about the prospects for liberal state-building in challenging environments. According to Marten, warlords themselves usually lack popular legitimacy, and central governments are unlikely to succeed in displacing those warlords using liberal means. Political leaders who must respond to popular opposition will find it more challenging to displace warlords once they are engaged. Either way, the message is not good for those who would advocate for liberal state-building as a national or international project.   Marten suggests a way out of this quandary through her argument that engagement with warlords is often a choice, not a necessity, but if it is, in fact, less of a choice than she suggests (and as I have argued above) then the prognosis for liberal state-building becomes even more dire. If extremely weak states—the failing and failed states that have been so central over the last decade—exist, then it is unlikely that they will be able to survive without some engagement with warlords. If this is the case, then it becomes hard to imagine how these states can transition out of a reliance on warlords into more genuine and legitimate political entities.

None of the above criticisms fundamentally undermines what is a very fine and very important book. Decades from now, scholars of warlordism will undoubtedly still be turning to Marten’s Warlords as an essential work on the subject. In fact, it is a virtue of the book that it raises so many of the questions that I have suggested above. Future scholars will do well to answer these questions as social scientists continue to struggle with understanding the many manifestations of sovereignty in the contemporary world and as political leaders continue to confront these manifestations in the making of policy.

Review by Matthew Evangelista, Cornell University

According to Kimberly Marten, the inspiration to write a book about the relationship between warlords and states came when she was embedded with the Canadian Armed Forces in Afghanistan in 2004 (1). That is the first feature that distinguishes Marten’s book from the typical political-science monograph: Marten did not set out to ‘fill a gap’ in the theoretical literature but to investigate a topic of concern to real-world politics. The last couple of decades have witnessed a steady disciplining of the discipline of political science to the point where aspiring doctoral candidates – especially in the subfields of international relations and comparative politics — believe there is only one acceptable kind of dissertation: the one that adopts a ‘mixed-method’ approach.   Normally that entails a simple game-theoretic formal model (even the complicated ones are typically designated ‘simple’) and a statistical regression based on an already-available or slightly supplemented data set, or — for the more ambitious students — an original data set, painstakingly assembled over months or years. These quantitative methods are then decorated with an illustrative, historical ‘case study,’ that is usually reliant on secondary sources. The field has seen a certain backlash to such formulaic approaches to generating knowledge, with prominent scholars advocating a return to historical research, with a stress on the method of ‘process-tracing’ to evaluate competing causal mechanisms of explanation. An important goal of process-tracing is to identify evidence that would support or undermine explanations derived deductively from existing theories.   Recent work offers ‘best practices’ and advice for “’fficient’ process-tracing so as not to include more historical detail than necessary.[2]

To her credit, Professor Marten has not allowed the methodological fashions of the discipline to constrain either her choice of subject or approach to studying it. Not that Marten is incapable of designing and carrying out a study to investigate hypotheses derived deductively from theory, as her previous books attest, offering excellent examples of the genre. That approach would not, however, have served her present purpose. In this case, Marten was drawn to the topic of warlords by a concern for public policy – namely, the observation that the United States and other countries were becoming increasingly dependent on “individuals who control small pieces of territory using a combination of force and patronage” (3) – her definition of a warlord. Marten’s general question is “what is the relationship between warlords, sovereign states, stability, security, and peace” (2)? In particular, the author wants to understand the policy-relevant question as to whether relying on warlords makes sense if a major power’s goal is to bring long-term peace and security to a weak or failed state. Marten does not avoid theory. On the contrary, in Chapter 2 Marten addresses a crucial theoretical question derived from the work of Charles Tilly – whether present-day warlords could “simply expand their territorial control and trading interests, and become state-building kings” (20), following the model Tilly described for early modern Europe.[3] Marten’s answer is a firm no, and the book’s case studies reinforce the point: “Because of the universal state system, today’s warlords arise and exist inside states where they are at least tacitly protected by domestic or foreign state leaders” (21). She argues that since warlords are parasitic on states, they are not interested in state-building. Thus, external powers that depend on warlords to control foreign territory (e.g., the United States in Afghanistan and Iraq) or domestic leaders who rely on them to maintain order in ungovernable regions of their own states (Russia in Chechnya, Georgia in several regions, and Pakistan in the Federally Administered Tribal Areas) should not expect warlords to provide an enduring solution.

Aside from addressing this key theoretical question, the author wanted to investigate the specifics of relations between warlords and their patrons. Unlike the textbook political-science study, Marten did not choose cases in order to test specific hypotheses.   She chose them instead for particular features they exhibit that help answer important policy-relevant questions. Marten was curious, for example, to know whether the current U.S. military vogue for understanding tribes and traditional leadership structures (to the point of recruiting anthropologists) offered a promising route to winning hearts and minds or at least controlling territory. Thus, Marten chose as the topic of her first case study (chapter 3) — the Federally Administered Tribal Areas, or FATA, on Pakistan’s border with Afghanistan – to examine the relationship between external powers and local tribes. The British colonial authorities, Marten argues, rather than respecting tribal structures, instead “used patronage to disrupt traditional Pashtun tribal norms and create a hereditary class of armed local power brokers in the FATA” (33).   Pakistan codified the British practices into its constitution, denying the FATA the democratic institutions, however imperfect, that the rest of the country enjoyed. The effect was a poorly policed border controlled by “ungoverned warlords” who exploited international development assistance to further their corrupt rule rather than as a “means for building popular support for the state” (61). Neither the goals of the U.S.-led coalition in Afghanistan nor of Pakistan’s civilian and military authorities were served by such an arrangement.

By using an inductive approach – rather than searching only for evidence to confirm or disconfirm certain hypotheses – Marten turned up valuable insights that neither the standard political-science techniques nor best-practice, “efficient” process-tracing would have produced. The author discovered, for example, that during the oil boom of the 1970s many residents of the FATA migrated to the Persian Gulf to earn and remit money independently of the local warlords. If not for the recession of the early 1980s and the decline in oil prices, Marten surmises, such financial independence might have provided a way to undermine the warlords’ power. In the concluding chapter of lessons and hypotheses, Marten suggests that “providing incentives for labor migration to state-controlled territories” could strengthen state authority and legitimacy by showcasing “the advantages of sovereign control” (195).

Marten’s use of induction does not mean that she neglects questions of research design. Her second case study of post-Soviet Georgia, for example, uses Mill’s method of difference to compare the approaches of two leaders – Eduard Shevardnadze and Mikheil Saakashvili – to dealing with warlords in Ajara and Upper Kodori, rebellious regions supported by neighboring Russia.[4]   Whereas Shevardnadze sought to accommodate the warlords, his successor Saakashvili managed to undermine their power. He used “fine-grained intelligence about how the warlord networks worked and who the individual supporters of the warlords were” to “peel away families who had earlier supported the warlords, through promises of future political and economic cooperation” (65). This “natural experiment,” as Marten calls it, “is the only modern case where state leaders followed such starkly contrasting policies toward warlords on their territories” (65).

Marten’s fourth case of Chechnya “is a striking example of an otherwise sovereign state methodically choosing to create a warlord on its own territory” (103). The case is unique in the sense that Russian leader Vladimir Putin “consciously created and abetted” the barely thirty-year old Ramzan Kadyrov as the rebellious republic’s reigning warlord, when, according to the author, there was no need to do so (137). But at the same time the Chechen case fits “a global (and historical) pattern: states choose to cooperate with warlords as a low-cost method for achieving immediate, short-term security benefits, without concern for the long-term consequences of their decisions” (105).

This latter generalization applies in spades to Marten’s final case – the use by the United States of the so-called ‘Sons of Iraq’ to help battle the Islamist forces associated with al Qaeda who rushed to Iraq in the wake of the U.S. invasion of March 2003 and its attendant chaos. Marten’s study calls into doubt the wisdom of trying to “work with the tribes” and the false “belief that they form a legitimate traditional authority structure in any state that has witnessed empire or dictatorship” (185). Her prognosis for Iraq, as for states that are reliant on warlords in general, is poor.

Warlords is a thoroughly researched and effectively written study. The benefits of eschewing conventional political-science methods, however, do entail some drawbacks. Investigating recent and ongoing developments demands reliance on a wide range of sources, and Marten has been particularly resourceful in tracking down subjects to interview and in making use of visual sources, from photographs to You Tube videos. For her Georgian case, for example, she was able to make judgments about the relative degree of support for warlords versus state authorities by interviewing participants and observers in demonstrations during the Rose Revolution and by contacts with Georgian officials and opposition figures. But in the absence of hard evidence for some points, the nature of her subject matter makes Marten sometimes overly dependent on rumors, claims, and counterclaims. Her text is interspersed with phrases such as “there were vague hints,” “there is no way to verify these numbers” (98), and “until the archives…are opened at some future date” (185). A key element of her argument about Putin’s according Ramzan Kadyrov an independent source of income by designating Grozny an international airport depends, for example, on whether Kadyrov’s security forces were able to operate there without interference from federal troops or police (120) – something Marten was not able to ascertain. To some degree this methodological difficulty comes with the territory, so to speak.   It is hard to gain as much solid information about shadowy warlords as about political and military figures whose actions are conducted in broad daylight. This point actually reinforces one of Marten’s important findings: “Information about local patronage networks (and how they may be diverted or supplanted) is the key weapon for states wishing to topple warlords” (192).

Another limitation of Marten’s approach concerns her judgments about democracy. Because Marten is studying weak and failed states, none of them with robust democratic institutions, the author sometimes attributes findings to the nondemocratic nature of her subject regimes. For example, an important element of Mikheil Saakashvili’s effort to buy off supporters of the warlord ruler of Upper Kodori, according to Marten, was his ability to spend $10 million in construction projects in a region of at most 2,500 people — a program that “would probably not have gone forward in a country subject to genuine democratic oversight” (96). Yet countries with far stronger democratic institutions than Georgia’s have managed to make substantial transfer payments and/or grant unusual levels of autonomy to regions at risk of secession, with the dominant populations going along (with greater or lesser degrees of reluctance) for the sake of national unity.   Cases that come to mind are Québec in Canada and the autonomous regions of Trentino-Alto Adige/Südtirol, Valle d’Aosta, Friuli-Venezia Giulia, Sardinia, and Sicily in Italy. Finally, even democracies can strike unsavory deals that are largely hidden from public scrutiny, as much of U.S. behavior in Afghanistan attests.[5] Such criminal means are typically justified by the expected ends: a secure and peaceful state. Kimberly Marten’s excellent book makes a convincing case for how unrealistic such expectations are.

Review by William Reno, Northwestern University

Kimberly Marten asks why leaders of states deal with local strongmen, whom she calls “warlords”, when their interests would be so much better served if they concentrated on building stronger state institutions that were subservient to their authority. For a long time scholars and policy experts have been interested in why leaders make decisions that weaken their own states. Most explain these choices with a focus on risk; on how shortened time horizons cause leaders to compromise with strongmen, pit them against one another, and undermine coup-prone bureaucracies just to stay in power, no matter the long-term impact on the authority of the state.

Marten’s contribution to this research agenda lies in her focus on warlord strategies and the wider context in which state rulers pursue relations with warlords. In this well written and engaging book Marten explains how warlords develop their own brand of authority through patronage networks and monopolize as much of the exercise of coercion as they can. Her significant insight comes in her discovery of the close, often intimate relationship between warlords and state authority.   Warlords in their contemporary form would not exist without close ties to state authority. The real secret of warlord authority is not found primarily in local patronage networks and militias, though these are important. The foremost source of warlord authority lies in subversion of the state. Warlords often convince others that that they can claim some of the prerogatives of state sovereignty. Some do this through holding an official position. Others hold no formal position but act as the right hand of the president or are used in other ways to assert the personal influence of a ruler. These relationships exist even if the warlord’s realm is a no-go area for the formal elements of capital-based authority.

Marten’s geographical focus does not include sub-Saharan Africa, though some of her ideas are most applicable to that continent. Her investigation into the symbiotic relationships between state authority (or, more precisely in many instances, the personal authority of state rulers) and warlords points to the value of sovereign status to those who challenge its basic precepts. Global norms forbidding territorial conquest and that accord sovereign recognition even to states like Somalia without effective central governments provide Marten’s actors with license to arrange their domestic affairs in ways that do great harm to state institutions. These are all issues that have occupied the attentions of scholars of African politics.[6]

Marten’s analysis resembles that of Pierre Englebertin his book Africa: Unity, Sovereignty, and Sorrow.[7] Englebert explains why strongmen—Marten’s warlords—and most other people in Africa do not just set out to make new states, even though existing states in some instances seem to have little to offer them. They maintain this connection not out of nationalist sentiment but for the capacity of global recognition of sovereignty to provide access to what Englebert calls ’domestic legal command’.   This means that pretty much anybody that can claim even a tenuous relationship with the state can equip themselves with the prerogatives of that station. For Marten’s warlords, this means that they collect customs fees, issue licenses and so forth as representatives of the state, however tenuous the claim, rather than as a replacement for it. These actors need the pretext of sovereignty and the illusion of subordination to gain foreign acceptance of their presence. Connections to even the weakest of sovereigns give them the capacity to lay claim to more resources associated with ’domestic legal command’ and to milk more aid from the international community. This explains why most warlords refrain from direct challenges to the sovereignty of existing states, even when their authority overshadows that of the country’s nominal central government. To define themselves as separatists or conquerors would bring swift international rejection and a much more challenging resource environment.

This political relationship to the state sets the stage for the conditions that define Marten’s warlords. The foremost condition is that contemporary warlords do not set out to build a sustained political alternative to the state, and instead stand ready to switch sides in capital-based political struggles to maintain access to the fruits of sovereignty. Warlords prove to be extremely reluctant builders of bureaucratic institutions. They are loath to delegate authority to people who might use these institutions to build their own powerbases to challenge their bosses. This condition drives Marten’s warning to policy makers that compromises and deals with local warlords are obstacles to real state-building. The logic of these actors’ relations to central authority makes the construction of durable bureaucratic institutions inimical to their continued political survival.

A critical reader may wonder whether Marten’s warlords differ all that much from the rulers of many states. After all, many rulers intentionally undermine their own state institutions, opting instead to rule though personal networks of patronage in which warlords so often play an unruly part. These rulers exploit the prerogatives of sovereignty and international protection against conquest and extinction to wring as much money from foreign governments and international institutions as possible.

The real distinction between warlords and state rulers in Marten’s argument lies in the capacity for radical choice. Since state rulers stand at the apex of domestic sovereignty, they are in the best positions to remake their political systems. They have to make their own calculations about the risks of abandoning their patronage networks and opting for greater short-term risks that would be involved in pursuing more rewarding long-term benefits of state-building. The durability of sovereign recognition at least enables them to make this choice. For the warlord, such a choice would be interpreted as a separatist challenge—a system destabilizer–and would attract sanction. For the ruler of a state, such a choice is received as the welcome action of a real reformer—a system affirmer.

This distinction underlies Marten’s advice to policy makers. Deals with local warlords in places like Iraq and Afghanistan are understandable when there is little political will or resources for policies that would encounter the political instability that would come with direct challenges to warlord authority. But as her study of Georgia Republic shows, some rulers are willing to undertake this risk for greater future gains. The book’s most provocative case study applies this argument in Marten’s analysis of U.S. involvement in Iraq. She argues that the Anbar Awakening of 2006-07 that turned Sunni militias against Islamist insurgence in alliance with U.S. forces was no great victory. Instead it empowered warlords whose personal authority rested upon undermining institutions of the state. Thus U.S. support for these actors played a significant role in ensuring that subsequent Iraq governments would have to deal with these warlords rather than concentrate on building effective state institutions.

The Iraq case illustrates Marten’s more general finding that the efforts of warlords to engage in their own style of inter-’national’ relations is corrosive of state-building efforts. This kind of engagement also is a key element of U.S. counterinsurgency strategy to seek out the relevant local counterbalances to insurgent authority. Particularly in rural areas, the personal networks of the local strongman provide foreign forces with ready-made channels of access to the population to identify insurgents and mobilize a local force to battle them. But this is a trap, Marten argues, that undermines the long-term goals of bolstering a state that can provide services and protect citizens while supporting local strongmen and their state-destructive logic of authority.

Thus Marten provides the reader with an inside-out explanation of how warlords endure and why foreign intervention can sustain them and undermine the state-building project. She also tells us where real state-builders come from. These are the rulers who seize opportunities to revise the local basis of authority, disrupting personal networks of power and replacing them with bureaucratic institutions. Georgia shows us how dangerous—and by extension, how rare—these efforts really are, and how important it is to support them. More typical is the record of Pakistan’s government. Instead of grabbing opportunities to integrate its borderland Pashtun populations into a growing economy back in the 1960s and 1970s, rulers opted to contract out governance in these regions to local strongmen, the warlords who presently support Afghan insurgents and defy the state along Pakistan’s sensitive border.

The implications of Marten’s argument and her advice to policy makers in this most provocative and stimulating book give the reader much to think about. Are foreign actors uncommitted or unknowing when they make   deals with warlords? The book’s implicit assumption is that state-building is doable, an assumption that might not have been as clear to military strategists who faced violent radical religious insurgents in Iraq from 2004. Not addressing this problem in the short-term may have led to a far worse outcome than Iraq’s current crippled state-building project. Likewise, civilian policy makers and military planners debate whether the government of Ahmed Karzai in Afghanistan has the political will to engage in a state-building project as they weigh the prospects for deals with warlords as the U.S. prepares to withdraw its forces there. In sum, expediency that empowers warlords may be adopted for good reasons. In many instances, those who adopt these strategies have no illusions about the poor prospects that warlords present in terms of state-building.

Is state-building really doable if domestic and external actors possess adequate resources and political will, as appears to have been the case in Georgia? This is an important question, given Marten’s critiques of American policies in Iraq and in Afghanistan. Consideration of a negative case would shed more light on this element of Marten’s argument. Such cases appear in Sierra Leone and Liberia, two small countries on the West Coast of Africa that offer for the scholar’s consideration a wealth of warlords and their close networks with state authorities. Unlike Iraq and Afghanistan, the populations of these countries generally welcomed armed international intervention in the form of tens of thousands of soldiers and massive aid programs larger than annual gross domestic output and that assumed direct control of critical state agencies. Yet even in these seemingly easiest of cases, foreign actors back off from direct challenges to warlords and instead make tacit deals with them. Reformers emerged, but they failed to embark on the kind of frontal assault on warlord authority seen in Georgia. The sudden collapse of Mali’s government in 2012 and the occupation of more than half of that country by violent religious extremists followed two decades of intense international engagement with reformers in the capital. That such an experiment in state building can collapse in a day or two (though signs were apparent long before) raises questions about the particularity of the Georgia case and the generalizability of the experience.

Though Marten does not delve into colonial histories of accommodation with local strongmen and warlords, this history sheds some light on why states and international actors compromise. The 1840 Durham Report following uprisings in Canada convinced Imperial British authorities that the effort to turn all of their subjects in Québec into English speaking Anglicans were bound to fail, and justified turning over essential state services such as education to the Catholic Church. The 1857 Mutiny in India taught British officials that efforts to integrate Indians under the centralizing authority of new state institutions would produce explosive political turmoil. Thus it became standard practice, codified in Sir Frederick Lugard’s Dual Mandate in Tropical Africa, to counsel handing as much authority as possible to what were essentially local warlords in the interests of   maintaining order at the lowest possible political and fiscal expense. These policy makers in the past did this because they discovered that a lot of warlords were popular and a lot of state-building projects were not. People rely on warlords to protect them from the cultural onslaught of ‘globalization’ and the frightening uncertainty that comes with economic growth.

All of this is not to argue that Marten has gotten it wrong. The author’s argument about the essential relationship between states and warlords breaks new ground. Marten offers a new way of thinking about big questions such as the prospects of modern states and the varied legacies of efforts to reform them. The book also warns policy makers to think about the unintended consequences and to ponder more realistically the expected outcomes of their choices of political allies and how they distribute resources.

The real value of Marten’s book is that it provides a rare platform for the scholar and policy maker to think further about the big questions surrounding state-building efforts. A policy maker reading the book might conclude that rather than simply dumping as many resources as possible on a country (as in Liberia and Sierra Leone), the more sensible strategy would be to focus on elements of the state that would deny warlords the benefits of ’domestic legal command.’ This would strengthen the hand of politically committed reformers. This assistance would avoid direct encounters with warlord authority that would provoke societal alarm and reaction. Instead of directly confronting warlords it would bolster state capacities to collect taxes and pursue those who evade them. Boosting customs services, port authorities, checks against cyber-crime, drug trafficking and financial fraud would dry up the pond of resources that are most often at the center of warlord relationships with state authority. Building the authority of states in this relatively stealthy and non-confrontational way would mimic some of the features of the historical strategies of state-building through war; in particular, the concentration of coercion and surveillance in the hands of state authorities. This would be a stealthier and somewhat more pleasant centralization, and hopefully one that would deliver more direct benefits to citizens.

Author’s Response by Kimberly Marten, Barnard College and the Harriman Institute, Columbia University

Let me begin by expressing my deep gratitude to the roundtable participants for reading and critiquing my book. I am especially grateful that they approached this task with such care and rigor. Each of the participants is a top scholar in one or more of the areas that the book covers, and I have long admired each of them. I’m honored by their participation, and relieved that their assessments of the book are largely positive.

Many of the points they make could be the basis for a second book project, and I welcome the efforts of others (including younger scholars) who might choose to use such critical starting points as a basis for their own future research. I will address below the points that more than one reviewer made, or that I find most compelling.

A key methodological argument that both William Reno and David M. Edelstein raise is that the book could have included more “negative cases.” Both wanted me to include more cases where states or other actors had the option of working with warlords, but chose not to do so.   They argue that this would have helped demonstrate the generalizability of my findings about how President Mikheil Saakashvili was able to oust warlords in Ajara and Upper Kodori when President Eduard Shevardnadze did not do so.

Since the primary purpose of my book is to generate and construct hypotheses using inductive methods, though, the inclusion of negative cases is not as necessary as it would have been had my goal been to test hypotheses. There simply is no developed literature on the relationship between states and warlords, so I had no deductive hypotheses to test. That is why I choose to end the book by outlining the series of hypotheses that my cases generated. As Alexander L. George and Andrew Bennett argue in their magisterial work on case study methodology, in the early stages of a research program it can be appropriate to choose even a single heuristic case to identify potential causal paths leading to a particular outcome.[8] (Matthew Evangelista’s critique, in my view, underestimates the degree to which George and Bennett support using case-studies for inductive theory-building, and not just deductive theory-testing.)

Further, as I say in the book, to my knowledge there simply are not very many negative cases from which to choose. There is one that my book mentions briefly in passing: the Somaliland region of Somalia. The secondary literature argues that traditional authority structures were sufficiently strong and popular there that warlordism was successfully beaten back, even as it engulfed the rest of Somalia. Reno is therefore correct when he says that the underlying premise of my book is that state-building in the modern world is doable: Somaliland might be where it happens. In retrospect it might have been ideal to include a deep case-study chapter on Somaliland in the book, in order to generate additional hypotheses about how warlordism can be avoided. Somaliland’s history as a British colony would make a nice comparison to the book’s cases of Pakistan and Iraq. (Of course, including more case studies would have made the book even longer.)

I discuss a different negative case in an earlier article[9] (also mentioned in passing in my book), namely revolutionary China. The secondary literature makes clear that Mao Tse-tung, unlike Chiang Kai-shek, refused to compromise with the leftover Republican-era warlords he encountered.   He either defeated them militarily, or else (if they wished to join his new Red Army) isolated them from their militias and put them through strict ideological reeducation, rather than allowing himself to be co-opted by them. I made a conscious decision not to include the contrasting Mao and Chiang cases in my book when I recognized that I could not find adequate source material for a rigorous case-study without reading Chinese. This case could now be used to test some of my book’s arguments, however, and I hope someone who has Chinese language skills might decide to tackle it.

In my earlier article I analyze what it takes for a society to rise up and throw off warlordism—namely, a combination of strong economic interests that are harmed by the existing system, and a strong and different idea about how society should be governed. (I would guess that both of those elements are present in Somaliland as well.) Edelstein is correct that a strong idea, namely the importance of Georgian sovereignty and territorial integrity, was also a key factor in Saakashvili’s popularity and his ensuing triumph against Aslan Abashidze and Emzar Kvitsiani. Edelstein is right that I might have emphasized this ideational component more as a key element of political capacity that Saakashvili held. Shevardnadze didn’t have that kind of populist vision.

In every other way, though—certainly including intelligence resources and bureaucratic strength—Shevardnadze’s Georgia had equal state capacity to Saakashvili’s (and probably an even greater ability to draw on state support from Russia). As the book makes clear, U.S. diplomats remained mystified about why Shevardnadze refused to take action against Abashidze in particular, since Abashidze was never very popular. This mystery grew with the Rose Revolution.   Shevardnadze’s alliance with Abashidze at that time led to his own political demise, while less than a year later Saakashvili removed Abashidze with help from members of Shevardnadze’s own Russian state network. If Shevardnadze had chosen to remove Abashidze, he would almost certainly have succeeded. The Rose Revolution was not a prerequisite for change in Ajara.

It will be interesting to watch what happens in this case in the immediate future. Free and fair parliamentary elections in Georgia in October 2012 removed Saakashvili’s people from leadership positions in Ajara.[10] So far there are no credible claims that anyone wants the old Abashidze regime to return. But the bargain Saakashvili made with his replacements is now broken.

Edelstein argues that many states may not have the choice to avoid working with warlords, since they simply lack the capacity to do so. He argues that this limits the generalizability of the Georgian cases. It is important to remember, though, that the Chechnya chapter in my book highlights a case where a state did make a choice: Russian President Vladimir Putin created his very own warlord by choosing to elevate one local militia leader, Ramzan Kadyrov, in a fashion that was quite unexpected.   Similarly, the secondary literature indicates that over time various governments in Colombia have chosen either to cede territory to FARC warlords or to take military action against them, while various governments in Mexico have made somewhat similar choices in regard to local drug gangs on their territory. It is surely not just Georgia that has had the choice of how to deal with warlords. The Colombia and Mexico cases, too, could be used now to test some of the hypotheses I generated.

But I fully agree with Edelstein that the lessons of the Georgian cases are far from being universally applicable. As I point out in my book, only states that have good intelligence resources, implying a significant level of organizational capacity, are likely to succeed in overthrowing warlords at a relatively low cost in blood and treasure.   Some of my book’s closing arguments concern how states should act if they find themselves forced to interact with warlords, for example by foregoing the temptation to legalize the relationship and thereby further undermine their intelligence-gathering capacities.

Reno suggests that I might have looked at Liberia and Sierra Leone. I am a bit confused, though, about what Reno’s argument is here. External military intervention and civil war led to the eventual overthrow of some very brutal warlords in both places. It is not surprising to me that warlords could successfully be overthrown by military force, but that really wasn’t the question I was looking at in the book. To examine these cases would require an in-depth comparative analysis of when and where the international community and rebel groups choose (or have sufficient capacity) to take action against warlord-suffused states. That could fill an entire additional book. Yet rather than seeing these as cases where warlords were overthrown by force (which I think would be the standard view), Reno refers to them as cases where “foreign actors back off from direct challenges to warlords and instead make tacit deals with them.” If I am reading Reno correctly, then the outcomes he cites here—external states supporting warlords—are ones that the underlying arguments of my book would predict. Working with warlords is the default option that states tend to select. In fact in a previous book on peace enforcement operations, I argue that the liberal democratic states that drive much of today’s international intervention are cost-conscious and fickle. This makes them incapable of forcing political change on foreign societies because their domestic societies have other priorities.[11] While I do believe that state-building is possible in the modern world, I am not optimistic that foreigners will do it well.

An important point raised by each of the participants in one form or another concerns the legitimacy of the warlord in the view of the local population. Edelstein is correct that I avoid any mention of the concept of legitimacy in the book. That is intentional. I wrestled for a long time with the question of how one might measure legitimacy in warlord territory, and I gave up.

In part this is because the literature on political legitimacy is replete with debates about definitions. One could accept the argument of Charles Tilly that whoever controls a piece of territory by force wins the legitimacy of the population.[12] From a different (but not incompatible) perspective, one could take the anarchist viewpoint, presented by John A. Simmons, that all seeming legitimacy is an illusion, since people naturally give their allegiance to whichever order they know.[13] My personal preference lies with the constructivist argument that legitimacy stems from a sense of social obligation and deference to a “rightful” order.[14] Rightfulness, though, can’t be measured by observing a population’s actions. As I argued long ago (in a book on post-Soviet defense industry), in a stable political and social order it is impossible to determine whether people conform to a given set of norms because they accept them intrinsically and emotionally, or instead because they recognize that even odious norms must be followed in order to further their own material self-interests.[15] If people obey and verbally support whoever has the dominant militia and the strongest patronage network, it is not clear that the concept of “legitimacy” has much analytic meaning apart from the concept of brute power.

Reno argues that sometimes local populations prefer warlords to the states that are present, and I agree. As I note in my chapter on Iraq, this appears to have been the case for Sunnis after 2003, who understandably preferred the protection of warlord militias to the violent sectarianism and neglect of the Shia-dominated state.   One of my book’s conclusions is that in stubbornly sectarian societies, warlordism might be the best achievable outcome.

Finally on the legitimacy question, I think Matthew Evangelista confounds warlordism with rebellion in his brief discussion of the actions taken by democratic states to prevent secession. While warlords and their supporters may threaten to back out of an informal bargain with a state, such a threat is causally distinct from a popular referendum that supports ethnic secessionism in democratic states.   Secessionism in twenty-first century Quebec is not functionally equivalent to a loyalty switch by 2,500 people (a third to a half of whom have served in a state-supporting private militia) in Upper Kodori. Among other things, in the case of Quebec the payments and compromises reached by the Canadian government are subject to ongoing public scrutiny and debate, which is one of the reasons why the secessionist question remains on the Canadian agenda after so many decades. There is no quick and easy private solution to Quebec’s dilemmas.

We expect democratic governments to give public goods payments to local populations on a regular basis, whether to buy their acceptance of a particular constitutional order or to buy their votes. Evangelista is also correct that the U.S. has made private side-payments in foreign countries under the rubric of military necessity, including in Afghanistan, and indeed over the past several years we have seen repeated Congressional investigations about the wisdom of those choices. But when representatives of democratic governments give private side-payments out of state coffers to individuals at home to buy their political loyalty, and fail to allow those payments to be debated in a public forum, that is usually called “corruption.”

Finally, Evangelista correctly points out that the evidence I use in the book varies in type and quality. As he recognizes, when examining cases with ongoing policy relevance that are sensitive from both the security and political reputation standpoints, it is sometimes impossible to find evidence from publicly available sources that is conclusive, or even cohesive. I am therefore as honest as I can be about the quality of each piece of evidence I use, so that readers can judge for themselves whether my analysis of the case is warranted by the evidence.

This book took years to research. My search for evidence was exhaustive, and I sought the criticism of every expert I could find at every step along the way. I spent countless weeks fretting over how to write the narratives for each case properly, and often when I was in the midst of a case I felt overwhelmed by what I saw as my responsibility to find and tell the truth as I understood it. Researching the Iraq case was particularly wrenching, because my desire to tell the truth of what I discovered sometimes conflicted with my desire to honor the lives of all of the people who died or were seriously harmed there.

As Evangelista notes, I did not approach this project by asking myself which questions I could answer with a particular data-set. Instead I asked myself which questions were important to answer, and then did the best job I could of finding evidence to address them.   No case-study is ever final, as long as future historians continue to unearth new pieces of evidence to challenge current understandings. The best an author can hope to do is to move the debate forward—and I think these reviews indicate that I succeeded in doing so.


© Copyright 2013-2015 The Authors.
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivs 3.0 United States License.



[1] Charles Tilly, “Reflections on the History of European State-Making” in The Formation of National States in Western Europe, ed. Charles Tilly (Princeton: Princeton University Press, 1975).

[2] Andrew Bennett and Alexander L. George, Case Studies and Theory Development in the Social Sciences (Cambridge, MA: Belfer Center for Science and International Affairs, 2005); Andrew Bennett and Jeffrey T. Checkel, eds., Process Tracing in the Social Sciences: From Metaphor to Analytic Tool (forthcoming). I was invited to contribute a chapter to the latter volume.

[3] Charles Tilly, Coercion, Capital and European States: AD 990 – 1992 (Hoboken: Wiley-Blackwell, 1992).

[4] John Stuart Mill, A System of Logic, ed. J.M. Robinson (Toronto: University of Toronto Press, 1973), discussed with application to international politics in Stephen Van Evera, A Guide to Methods for Students of Political Science (Ithaca, NY: Cornell University Press, 1997).

[5] Rod Nordland, “Top Afghans Tied to ’90s Carnage, Researchers Say,” New York Times, 22 July 2012; Democracy Now, “Congressional Probe Reveals Cover-Up of ‘Auschwitz-Like’ Conditions at U.S.-Funded Afghan Hospital,” 1 August 2012,

[6] Robert Jackson, Quasi-states: Sovereignty, International Relations and the Third World, (New York: Cambridge University Press, 1990); Christopher Clapham, Africa and the International System: the Politics of State Survival, (New York: Cambridge University Press, 1996); Jeffrey Herbst, States and Power in Africa: Comparative Lessons in Authority and Control, (Princeton: Princeton University Press, 2000).

[7] Pierre Englebert, Africa: Unity, Sovereignty, and Sorrow, (Boulder: Lynne Rienner Publishers, 2009).

[8] Alexander L. George and Andrew Bennett, Case Studies and Theory Development in the Social Sciences (Cambridge, MA: MIT Press, 2005), pp. 23-4.

[9] Kimberly Marten, “Warlordism in Comparative Perspective,” International Security 31, no. 3 (Winter 2006/7): 41-73.

[10] “Georgian Dream Winning Majority in Adjara’s Supreme Council,” Civil Georgia (Tbilisi) Oct. 3, 2012, available at .

[11] Kimberly Marten, Enforcing the Peace: Learning from the Imperial Past (New York: Columbia University Press, 2004).

[12] Charles Tilly, “War Making and State Making as Organized Crime,” in Bringing the State Back In, ed. Peter Evans, Dietrich Rueschemeyer, and Theda Skocpol (Cambridge: Cambridge University Press, 1985), pp. 169-86.

[13]John A. Simmons, “Political Obligation and Authority,” in The Blackwell Guide to Social and Political Philosophy, ed. Robert L. Simon ( Hoboken: Wiley-Blackwell, 2002), pp. 17-37.

[14] Michael Barnett and Martha Finnemore, Rules for the World: International Organizations in Global Politics (Ithaca: Cornell University Press, 2004); Deborah D. Avant, Martha Finnemore, and Susan K. Sell, “Who Governs the Globe?” in Avant, Finnemore, and Sell, Who Governs the Globe? (New York: Cambridge, 2010), pp. 10-11.

[15] Kimberly Marten Zisk, Weapons, Culture, and Self-Interest: Soviet Defense Managers in the New Russia (New York: Columbia University Press, 1997).