Where do effective military and police institutions come from in a society that is not already based on the rule of law? In particular, can informal militias based on patron/client relations be successfully reformed or integrated into professional and effective state security institutions? We do not have good answers to these questions. Yet the United States and its allies are wrestling with them daily in many locations around the globe. My goal in this article is to examine what we do know about historical and recent situations that to some degree mirror these current challenges, and to draw out some unexpected practical suggestions about what might work on the ground.
These questions originally grew out of my research on warlords. Warlords are individuals who control small pieces of territory through a combination of force and patronage, acting in defiance of genuine state sovereignty but with the collusion of weak states and their leaders.1 The relationship of warlords and their informal militias to state actors is bargained and based on personal ties. Warlord militias are not implacably hostile to the state or resentful of de jure state sovereignty over the territory where they operate. In most cases, state leaders have actually informally granted them de facto control over particular territories. This situation creates obvious challenges for internationally supported security sector reform efforts in places such as Afghanistan.2 But these questions have relevance beyond cases of warlordism, too. Rebel or paramilitary forces opposed to the current state can also be based on patronage ties, reflecting a commander's ability to take care of followers by sharing the spoils with them rather than on professional discipline and adherence to a mission or ideology.3
Such questions raise a conundrum for the United States and its allies in many postconflict areas where key interests are at stake. As a crucial component of peace accords in many parts of the world, militias that are based as much on personal ties and the political economy of a spoils system as on common ideological goals have been integrated into new state security forces, with the United States and its allies providing training and support. The prime consideration for policymakers has usually been to stop the bloodshed and achieve some measure of stability in new or reconstituted states. The successes of these stabilizing efforts should not be underestimated. There has been little focus, however, on the long-term security implications for the United States when it chooses to support patronage-based security institutions under a new guise.
For example, in Afghanistan the Local Police Initiative relies in part on former mujahideen commanders (at least some of whom fit the definition of warlords) for recruitment through personal connections.4 In Iraq, many of the Awakening and Sons of Iraq units that were slated for integration into state police and army forces were tribal or neighborhood groups, connected through family or other personal ties, who were known to have engaged in organized criminal activity in the past. In Kosovo, family and clan relationships underpinned many Kosovo Liberation Army units, which have now been quietly integrated into the new state security forces after what amounted to a temporary decade of official disarmament. In Somalia, the United States and its Ethiopian allies have often cooperated with former warlords and their chosen armed contingents.
A vivid illustration of this conundrum is found in the Egyptian military in 2011, which has taken a leading role in the post-Mubarak political transition following popular revolt against the authoritarian regime. Long trained and supported by the United States, Egyptian forces have been lauded for their combat skills, respect for civilian control, and efforts at civilian casualty avoidance. Egyptian protestors were eager to maintain good relations with the military, seeing the institution as friendly and a potential bulwark against the brutal Interior Ministry police. Yet the Egyptian military is known to practice patronage. Analyst Lisa Anderson notes that the military leadership is "deeply interwoven into the domestic economy" and "largely hostile to economic liberalization and private-sector growth."5 It remains unclear what this military patronage system bodes for the future of Egyptian development and for Egyptian popular perceptions about what has been a critical U.S. alliance in the volatile Middle East. This example suggests patronage is a factor the United States should consider when designing its foreign military assistance policies.
Definitions and Implications
In patron/client systems, both security and economic advancement are determined by how well one is personally connected to those who control the exercise of force. People in the inner social circle surrounding powerbrokers are well protected, and they and their friends have easy access to political and economic resources. Patronage ties may lead to hiring and promotion as well as opportunities for graft and corruption that are denied to those outside the network. Those who are not favored by family or other social connections must simply live as best they can, on the good graces of those who control the territory. For example, if they try to start a business that is not supported by the powerbrokers, they may have a hard time protecting it from thieves or arson.
Afghan Uniform Police stand guard as officials meet with elders about local initiative
In these systems, trust of other people depends on personal relationships, not on faith in abstract institutions such as the law. This means that strangers are by definition threatening since they belong to different and potentially competing patronage networks. Thus, laws and other institutions of the state are rather meaningless; they are shells controlled by in-groups. Corruption of state resources—not rule of law—is the socially accepted norm.
As long as commanders in patron/client systems retain the loyalty of their forces, patronage-based militias (and what Samuel P. Huntington famously called "praetorian" state security institutions6) can win battles and keep and expand territory. Francis Fukuyama points out that China, for example, has wrestled for centuries over whether familial patron/client ties or professional bureaucratic institutions would be the basis for state-building.7 The Chinese state endured through all of these struggles, which he believes remain unresolved to this day.
Fukuyama argues nonetheless that the Chinese state was weakened whenever familial rule won out. An examination of patronage systems in theory and practice explains why. Distrust of outsiders is endemic, making large-scale coordination difficult and violent infighting between competing in-groups likely. Fukuyama notes that familial rule often led authority to devolve to the local level in Chinese history. Corruption makes security institutions expensive to maintain, and implies that their equipment, logistics, and even purported personnel numbers will be unreliable, given the incentives to fudge the data for personal gain. As we see in Egypt, by their nature these institutions set favored in-groups apart from disfavored out-groups, potentially creating long-term resentments (and hence instability) both within the forces and the surrounding population. Such characteristics cannot form a solid basis for reliable security institutions and state-building over the long term.
The lack of attention to this problem by U.S. policymakers is not surprising, since these questions also fall into a gap in the literature on state-building. There is a large academic literature from sociology and economics that describes why rule-based (as opposed to personal patronage-based) economic institutions should evolve over time. The institutionalization of abstract rules that apply to everybody should serve the interests of elite actors. Elites enmeshed in traditional patron/client exchange relationships should find that their interests are better served by being more inclusive in their trading and investment decisions since this expands the scope and scale of useful exchanges. The inclusion of strangers in economic relationships demands a high level of confidence in legal institutions that can enforce bargains and contracts.8 Buyers and sellers should value the reduction in uncertainty that legal institutions provide for their commercial interactions. Third-party enforcement of legal bargains allows them to decrease the costs of business transactions, especially when dealing with potentially unscrupulous strangers.9 State leaders should come to see it being in their interests to enact property rights, rather than ruling by capricious or arbitrary fiat, since the confidence this gives property holders in the future value of their investments will lead to revenue maximization (useful both for paying off internal rivals and fighting foreign wars).10 In turn, when wide-ranging, inclusive bargains can be enforced through institutional legitimacy and general respect for law, rather than the use of actual force, it frees state resources for wealth generation. Institutionalization of economic law thereby contributes in a virtuous circle to the future bargaining power of states and state leaders on both the domestic and international level.11 The literature agrees that this helped lead to the triumph of the Western European model of statehood since reliable legal institutions allowed such states to achieve military and economic dominance.
Yet there is a gap between the economic institutionalization literature and the notion of the institutionalization of security provision. We know full well that many states in the world continue to operate by patronage in spite of these supposed incentives for change. Who does the enforcing of these new inclusive bargains and depersonalized laws in a system that is not already based on abstract legality? How are security forces found who are willing not to be bought off by the old patron/client networks?
All human interaction originally began at the level of families, clans, and tribes.12 Security and defense of the group from predators was provided by those whom one knew personally. In a stateless social environment, it is natural for security institutions to practice patron/client favoritism and out-group predation since both trust and profit are based on direct knowledge of another person's history and reputation. Many organized crime groups in areas suffering from state weakness continue to operate through such patron/client relationships today.13
Yet at some point effective military and police forces, including those located in liberal democracies, stopped practicing patronage or at least severely limited its practice (one can imagine, for example, that nepotism in police hiring might still be common, but with forces for the most part practicing rule-based, rather than family-based, enforcement of law). How can security forces that are undergoing the process of state-building or state reconstruction come to be made up of individual members who follow abstract, institutionalized rules instead of remaining loyal to patron/client cleavages? How, while undergoing that process, can they maintain their ability to defend themselves and their populations against those whose self-interest is harmed by the end of patronage? In other words, how can rules be made stronger than personal social ties in situations where legalism is not already the norm and security is tenuous?
Our comparative historical knowledge about these endeavors is weak. In the rest of this article, I review some of the existing literature relating to these questions, identifying crucial issues for policymakers to consider, even though most of the literature does not address these questions directly.
Literature on Demobilization
There is a large recent literature on the demobilization of fighting forces after civil wars.14 By definition, the concept of demobilization implies that rebels are disarmed—nonstate militias are broken down and disbanded. There is no effort in these cases to integrate informal militias into the new security forces, which are instead formed from scratch. For example, in postwar Liberia, officers of Charles Taylor's army and police forces, as well as anyone known to have committed any crimes or human rights abuses, were explicitly excluded from the new security agencies.15 In postwar El Salvador, a national police force was created de novo, state military forces were reduced and reformed, and rebel forces were disarmed and became a successful civilian political party.16 Getting rid of the old security forces might seem a straightforward way to eliminate their patronage behavior.
Indeed, forced or negotiated demobilization of nonstate militias has been a regular occurrence throughout history. In Europe, it dates at least from the rise of European nationalism and mass armies under Napoleon.17 In Japan, it occurred at the dawn of the Meiji era in the late 19th century when samurai lords were stripped of their military functions to fashion a new state that could compete with the West.18 In China, it happened in the 1930s and 1940s, with the triumph of Mao Zedong's Red Army out of the ashes of Republican-era warlordism and the eventual defeat of Japanese efforts at domination.19 The newer literature on the concept never cites or explores this historical analysis, and that is unfortunate; the problem of how to successfully demobilize a militia is not a new one.
The findings of this older literature lead to two potentially useful policy conclusions. First, historically leaders were able to replace patron/client security systems with effective professional armies when states faced a strong external threat. In Europe, China, and Japan, demobilization of old militias happened when the state needed an effective army to ward off foreign invasion. (Fukuyama notes that militaries often have no means to survive in combat unless they become meritocracies; he believes that this causes military organizations to lead civilian bureaucracies in the speed and depth of their professionalization.20) This may mean, though, that such a transition might have trouble succeeding today. Most states simply do not face a threat to their existence. Scholars have long noted that the absence of external threats has enabled the survival of weak states in postcolonial Africa.21 It also may mean that there is little external pressure forcing patron/client-based security systems to transform themselves.
Afghan Uniform Police knock down poppy plants in Zabul Province
ISAF (Brian Wagner)
But perhaps external conflict is just an enabling factor, and not a necessary condition, for change. A second factor present in each of these historical cases was an ideological transformation of broader society that accompanied the response to the external threat. State (or proto-state) efforts to establish a professional army were bolstered by popular acceptance of (and even fervor on behalf of) new nationalist belief systems that united people in identifying with a greater geographic space. In China, this was enhanced when Mao's vision of communist equality won over the peasantry. Today, nationalism per se is not a new ideology. Yet an ideology of individual human rights, growing out of the western Enlightenment, fundamentally challenges the collectivism and familialism of patron/client societies. This seems to be playing a role in the current revolutionary fervor suffusing the Middle East. It remains unclear whether such a liberal ideology can motivate revolutionary change on the scale that nationalism or communism did. Nonetheless, these findings suggest that the promotion of literacy, education, and communications and transportation infrastructure (to enable the transmission of new ideas) may be important mechanisms for replacing patron/client-based systems today.
Perhaps the major reason the modern literature on demobilization does not refer to these historical cases is that today's preferred demobilization method does not center simply on destroying and disarming the old militias. Instead, with the support of the United Nations (UN) and other members of the international community, the process is just one component of demobilization, disarmament, and reintegration (DDR). Militia members are given new skills, new jobs, psychological and career counseling, and community support in hopes that this will discourage them from taking up arms again.
It is on these (usually negotiated) DDR processes that the current literature focuses, cautioning that demobilization is tricky and prone to failure. Those who are used to the prestige and power that comes from wielding guns may be loathe to give them up and may fear for their own safety if they lose the ability to defend themselves against rivals or retribution attacks. Militias may become spoilers of peace processes,22 or they may turn from warfighting to organized crime and gangster activity.
Mark Knight and Alpaslan Özerdem warn that the possibility of failure is magnified when nonstate militias are cantoned together in groups for reeducation and processing during the demobilization period. Cantonment isolates the militias from the rest of society and can potentially reinforce existing command structures.23 Macartan Humphreys and Jeremy Weinstein similarly find that separating combatants from their former militia comrades was the most important part of the successful DDR process in Sierra Leone. Such "delinking" of former militia members, especially when it came to economic opportunity, was a much more important indicator of true reintegration than participation in official UN DDR programs.24 In the same vein, Sarah Zukerman Daly shows that densely networked former combatants in Colombia who hail from (and return to) the same locality can easily reconstitute themselves into paramilitary organizations when power balances shift.25 When groups retain their predemobilization social structures, the influence of old commanders and operational mindsets remains stronger.
While these findings are usually presented in relation to the question of whether civil war may break out again, they can also be applied to the question of what leads to the successful creation of professional state security forces out of a patronage-based past. They imply that when old militias are kept together as units rather than dispersed, they are less likely to be successfully integrated into well institutionalized state security forces and more apt to maintain previous patron/client ties. This conclusion might seem obvious, but it is striking how often it has been ignored by policymakers. Attempts to integrate by unit, under existing command, are frequent. For example, in Republican-era China in the 1920s and 1930s, Nationalist Army leader Chiang Kai-shek tried to integrate warlord units into his forces. He learned to his chagrin that it resulted in corruption, theft of resources, and fragmented command in battles against Mao's superior Red Army.26 More recently in post-Dayton Bosnia, army brigades (later switched to battalions of 1,200 soldiers each inside brigades made up of different ethnic battalions) have been drawn from former ethnic militias. These battalions seldom have much substantive contact or cooperation across ethnicities, instead reinforcing old networks. A similar arrangement has been reached for separate ethnic entity policing in Bosnia.27 Even more recently, in Al Anbar Province in Iraq, preexisting Sunni tribal militias were recognized as state-sponsored local police forces under the state-supported Awakening program. They refused an alternative proposal from Baghdad that would have dispersed their members throughout Iraq, arguing that dispersion would subject them to sectarian Shia attack and that they wished to protect their own neighborhoods instead.28 The evidence from the demobilization literature suggests that in these cases, patron/client ties will continue to dominate security force relations and prevent the true professionalization of forces. Short-term stability may undercut long-term state-building and state security.
Literature on Rebel Integration after Civil War
There is an emerging literature that focuses on the integration of militias into state military forces after civil war rather than DDR.29 It asks how rebel groups with clear territorial aims should be treated after civil wars to ensure that war does not break out again. Its authors concentrate on issues of power-sharing and trust-building between former enemies rather than examining the question of patronage versus professionalism within the new forces. The argument behind much of this literature, for example in work by Caroline Hartzell and Matthew Hoddie, is that power-sharing is a postwar confidence-building measure.30
This literature concentrates on what it claims are some notable successes of rebel integration into state security forces. Yet these claims are problematic for two reasons. First, the literature does not examine variation in the long-term results of integration efforts. Its goal instead is to assert that integration can work, and to give examples where that has been the case in the immediate term. Unlike some other authors, Stephen Burgess does look at cases of integration failure rather than merely concentrating on the success stories. However, he defines failure as the collapse of the peace agreements that support the integration process.31 He does not examine in any depth variation in what happens to state security institutions after the integration process has occurred. Like Hartzell and Hoddie, his focus is fundamentally on what makes power-sharing arrangements succeed at a political level.
The second and related problem is that success in this literature is defined in such a way that it does not reveal anything about the reorientation of security forces from patronage toward professionalism. While the success or failure of power-sharing agreements can establish whether civil war ends, such agreements alone cannot determine whether a newly integrated military or police force will be characterized by rule-abiding integrity or instead by in-group favoritism and corruption.
Indeed for some of these authors, integration success merely implies that the process of hiring militia members into state security forces proceeds smoothly. In this view, success is achieved when a significant number of militia members are offered jobs and accept them, and civil war between the rebel group and the state does not break out again, no matter how ineffective the new security forces are.32 For example, Florence Gaub argues that the primary benefit of integration in both Lebanon and Bosnia was "symbolic," serving as an indicator of "the return of the rule of law and the state in all its functions" after civil war (even when that rule of law and state functioning are essentially toothless and old militia ties remain strong).33 The Mindanao experience in the Philippines is categorized as a success by some authors. Yet a case study that lauds success in Mindanao also hints that security force integration resulted in command and control difficulties and financial mismanagement.34 Burgess calls the Zimbabwe integration experience a success, even though "fighting between former . . . factions marred the integration process" and "in spite of corruption and decay."35 Uganda's integration experience is similarly termed a success by Monica Duffy Toft.36 In Uganda's case, even some members of the notorious child-snatching Lord's Resistance Army, which was blamed for thousands of war crimes, have recently been turned toward cooperating with the state.37 Yet a recent news report laments that the Ugandan army is "bolstered by corrupt spoils," and cites a statement by British scholar Phil Clark that the Ugandan leadership rules through "the fear of violent military crackdowns."38 This finding is especially disheartening since Jeremy Weinstein had earlier described the formation of the National Liberation Army (the rebel group whose victory in warfare led them to become the new Ugandan state) as operating under a population-centric ideology based on strict discipline and respect for civilians.39
Burgess defines success in more detail than many other authors in this literature. In his view, success is measured by the survival of the new regime in a high-threat environment, indicating that the security forces are able to protect the state, and sometimes by the ability of the newly integrated forces to engage in foreign intervention or peace enforcement processes without being routed.40 This fits the general definition of military effectiveness put forward by Stephen Biddle as well, who in a recent literature review noted that the "ultimate dependent variable of interest for effectiveness is the outcome of combat."41
Yet the world is replete with examples of military and police organizations that have managed effective security states (at least temporarily) or survived opposing forces by cowing the public while remaining corrupt and undisciplined. If this is how success is defined, then one might argue that Ramzan Kadyrov of the Russian republic of Chechnya has succeeded in integrating rebel forces into his autonomous militia (now recognized by the state as the local branch of Interior Ministry forces) through a campaign of well targeted and selective terror.42 Indeed, Russian authorities do hold Kadyrov up as a success story since outright civil war in Chechnya is over. Yet Kadyrov's successes are fundamentally centered on a brutal form of personal patronage—and few objective analysts believe that the deep security problems of Chechnya and its relations with the Russian state have been permanently solved.
The definition of success in the existing literature is further complicated by the fact that many nonstate militias have links to foreign actors, including states, multinational companies, and nongovernmental organizations.43 Warlords are particularly susceptible to the lure of middle-man status, taking on patron/client roles between states and foreign actors.44 Links with foreign benefactors who evade state control—whether they provide weapons, investment, training, or simply political advice and friendship—need to be broken to ensure the long-term reliability of successful integration into professional home-state security forces. The existing integration literature remains largely silent on that topic.
For these reasons, it is not clear that the literature on the integration of militias into state security forces offers any useful conclusions for the questions presented here, even though integration of informal militias into state security institutions is a fundamental subject of this study. Integration, in many recent cases, does not seem to be associated with professional transformation.
Literature on the Motives of Rebel Forces
There is a well established literature examining the motives of rebel forces.45 It offers a range of competing explanations for why rebels fight. These explanations range from the authority of family and community norms to the importance of personal values and emotions, and include a focus on the power of economic and political incentives that range from the individual to the collective and from the structural to the strategic.
The variation in explanations offered by this literature makes it hard to draw any definitive conclusions from it. But part of this literature deals with the fact of indiscipline and corruption within certain rebel movements but not others, and two useful conclusions can serve as the basis for further policy advice. First, Weinstein argues that rebel organizations blessed with access to economic endowments (oil, diamonds, or external financing, to name a few) will by nature end up attracting recruits based on the possibility of looting and other immediate access to wealth. He believes that rebel organizations not blessed with this access have to rely on what he calls "social endowments" for their recruiting power instead. He argues that they will develop norms of group solidarity, focused on the achievement of a far-off political or ideological goal. He also believes that these resource-deprived rebel groups are more likely to treat civilians well since they need local support and trust in the absence of alternative sources of financing. In other words, rebel groups based on social endowments are more likely to follow the rule-based patterns of professional armies. Rebels attracted by economic resources will be more prone to fragmentation and more apt to engage in indiscriminate violence against civilians, both markers of indiscipline.46
This can lead to a related conclusion, that militias whose members are denied access to private economic endowments after integration (for example, lootable minerals or narcotics) will be more successfully integrated into well institutionalized military forces than will those who gain or retain access to those endowments. Those who retain access to lootable resources will be more likely to maintain their patronage behavior. To speak plainly, this might mean that success in building professional security institutions in Afghanistan depends on eliminating the illegal poppy trade. Otherwise, local militias that are now being integrated into state security forces have a strong incentive to continue illegal patron/client relations on the side, thereby preventing the institutionalization of professionalism. Current U.S. and North Atlantic Treaty Organization policy that avoids poppy eradication in Afghanistan may be wrongheaded, as it sacrifices long-term success in building reliable state security institutions for the short-term benefit of hearts and minds operations with local poppy farmers.
Paul Kenny's work on rebels and lootable resources takes a less structural and deterministic approach than Weinstein. While Kenny agrees that resource extraction is tied to force cohesion, he focuses on rebel leadership strategies in dealing with those resources, not on the question of whether fixed background conditions influence cohesion and disintegration. When leaders of rebel groups are "forced to be self-financing" or choose to be "dedicated primarily to resource extraction," in Kenny's view, cohesion will fall. In that sense, he agrees with Weinstein. But he goes on to argue that experiences of "shared sacrifice" on behalf of a common goal (for example, rebels serving prison terms together on behalf of their cause, or accepting tattoos that permanently mark their rebel status and thereby prevent them from denying it if they are captured) will foster socialization into the organization and lead to cohesion.47
If this argument holds true postintegration, it means that nonstate militias whose members undergo shared sacrifices alongside other state security forces will be better professionally integrated than those who do not share that experience. This provides an alternative explanation for the success of both the Zimbabwean and postgenocide Rwandan integrations that are described by Burgess. He argues that a mark of the success of the professional integration process in Zimbabwe was that the new forces fought well in Mozambique and the Democratic Republic of the Congo (DRC) in the 1980s and early 1990s.48 In postgenocide Rwanda, he argues that a similar mark of success was that former Hutu militia members were effective at fighting Hutu extremists in the DRC in the mid to late 1990s and 2000s.49 Kenny's argument suggests, however, that the causal pathway may be reversed: shared sacrifices that these forces underwent during combat in the DRC might have helped both countries achieve the successful integration of former enemies. In other words, sending newly formed or reformed militaries on difficult peace enforcement operations abroad may help them become more effective and stronger institutions.
Some analysts throw up their hands and declare the problem of security sector professionalization insoluble in patron/client systems. Dipali Mukhopadhyay, for example, in her brilliant study of how warlordism has continued in Afghanistan under Hamid Karzai, argues that while warlords continue to exact "undeniable costs" on the citizenry, the patronage bargains they have reached with the Afghan state "may represent the best dividend to be expected" in the absence of "credible, pre-existing domestic security institutions." She believes that security sector reform in Afghanistan "reflects an assumption of formal institutional capacity on the part of the state that simply does not exist."50 Citing Charles Tilly's work, she reminds us that state-building in Europe was messy, violent, and based on informal bargains, too. Maybe the continuation of patronage systems, for the sake of immediate stability, is the best we can do.
Certainly the problem of overcoming patronage systems is a difficult and long-term endeavor for those seeking to build or assist the creation of new security institutions. This article nonetheless suggests that the history of and literature on state-building and security sector integration can be mined for advice. There may be creative ways of thinking about the design and policy environment of new security forces in areas of the world where patron/client relations have been the norm, resulting in small steps forward on what is undoubtedly a long path to eventual professionalization. PRISM