Behavioral Patterns among (Violent) Non-State Actors: A Study of Complementary Governance

This article is part of a multi-year study of governance structures in the midst of insecurity and organized crime in fragile sub-state regions where, in the absence of a strong state, non-state actors (like insurgents, traffickers and tribal warlords) engage in political and socioeconomic governance. Building on our prior work on West Africa and the Afghanistan-Pakistan tribal belt, this paper focuses on the Andean borderlands, drawing on recent fieldwork in Colombia, Ecuador and Venezuela. We explore patterns of behavior in which competition among violent non-state actors is not the norm. Instead, several instances were found in which violent non-state actors work collaboratively or have tacit non-interference agreements to provide public goods through arrangements we characterize as ‘complementary governance.’ We therefore argue that, to understand how illicit authority emerges, it is not sufficient to consider one armed non-state actor in isolation or in a dichotomy to the state. As we contend, we have to take into account the complex connections and interactions among different (violent) non-state structures. Moving beyond state versus non-state governance to governance that is constitutive of various non-state groups, the perspective put forward in this article aims to enrich the current debate on governance and security.

Concern about state weakness has risen to an unprecedented level, featuring prominently in national security strategy documents, the academic community, and major public awareness efforts like the annual 'Failed State Index' produced by Foreign Policy magazine and the Fund for Peace (2014). 1 Terms and phrases like 'governance voids, ' 'ungoverned spaces,' or 'lawless areas' have been increasingly used in reference to many of the world's most troubling security challenges, from brutal war scenarios in Africa to soaring homicide rates in Latin America and bloody unrest in the Arab world (Koonings & Krujit 1999;Koonings & Kruijt 2004;Clunan & Trinkunas 2010;Rabasa et al. 2007;Stigall 2013). Scholars and policymakers have frequently cited the incapacity of governments to govern (parts of) their country as a major challenge that has implications for both local human security and the international security environment.
A central tenet within these discussions is that when states are weak, or absent Idler and Forest: Behavioral Patterns among (Violent) Non-State Actors Art. 2, page 2 of 19 altogether, a power vacuum can emerge, into which other actors will exert their own authority. For example, the Shiite militia group Hizballah reigns supreme throughout southern Lebanon (Levitt 2013); Tuareg and Berabiche groups exercise power and authority in various parts of the vast Sahel region of Africa (Kennedy-Boudali 2009: 4); on the island of Mindanao in the Philippines, Muslim inhabitants commonly known as Moros or Bangamoro (the Moro nation) have contested the authority of the Manila government and its religious and cultural influence since the Spanish colonial period (Rabasa et al. 2007: 4); and transnational criminal networks have significant power throughout the tri-border area of South America (where Argentina, Paraguay, and Brazil meet). Widespread concerns about the security implications of these developments have led to a significant increase over the past two decades in research on the intersection of governance and violent non-state actors (VNSAs) 2 such as rebel or paramilitary groups (Wood 2003;Kalyvas, Shapiro & Masoud 2008;Arjona 2010;Mampilly 2011;Staniland 2012).
Non-state actors -violent or otherwisewho have power over a local populace often play by a different set of rules than the formal governments of nation-states. Trust is established not by a legal system or formal contract between a leader and those governed, but by informal systems of traditional customs and moral codes. For example, in countries like Nigeria and Senegal nonstate entities engaging in some semblance of governance may include traditional clan and tribal leaders entrenched within deeply established patronage networks. In certain parts of northern Nigeria religious leaders (like the Sultan of Sokoto and his various local Emirs) wield considerable authority among Hausa and Fulani communities. In many parts of south Lebanon, Hizballah's permission is required for anything from the acquisition of building permits to the provision of social services. On every continent scholars have identified non-state actors who provide a functioning security and intelligence apparatus, an infrastructure for commerce and transportation, and even a local customs-based or religious mediation system (such as tribal councils or Sharia courts) for resolving disputes -none of which are necessarily controlled or even sanctioned by the nation-state. Unlike the image that comes to mind when using the term 'ungoverned spaces,' there is actually some sense of order here, but the nation-state is not considered the primary or effective authority that maintains that order.
Typically, these non-state actors are characterized by the research and policy communities as rational actors who make strategic decisions in pursuit of their own self-interests. From this perspective, a common explanation for non-state actors providing public goods and services is that they do so in exchange for acceptance of their authority (Forest 2010;Forest 2011;Idler & Forest 2012). Territorial and political control is seen as a central motivator of non-state actors' decision-making and behavior. Further, this leads many observers to assume that the political and socioeconomic environment of weak states is one of inherent competition for power and control among non-state actors (including VNSAs), with a feeble state government trying in vain to exert some level of control or influence over the population (Jackson 1990;Dasgupta 2004;Shultz, Farah & Lochard 2004;Eizenstat, Porter & Weinstein 2005;Grynkewich 2008;Hesterman 2013;Miklaucic & Brewer 2013;Bunker 2013;Fund for Peace & Foreign Policy 2014). 3 This article provides an alternative perspective by asking: How do (violent) nonstate actors relate to each other in ways that provide local public goods not otherwise provided by the state? Drawing on recent field research, we describe patterns of behavior in which competition among VNSAs is not the norm. Instead, from an extensive case study of the Andean border region, several instances were found in which VNSAs actually work collaboratively or have tacit non-interference agreements to provide public goods through arrangements we characterize as ' complementary governance. ' Our understanding of complementary governance is drawn from research on how authority is distributed between governments and non-violent non-state actors in various countries throughout the world. Examples of this range from Pashtun tribal leaders in Afghanistan to Tuaregs in northern Africa and the nomadic Bedouins in the Middle East. A prominent case is seen in Nigeria, where a complex web of religious leaders and local traditional rulers exercise a great deal of authority and power. Centuries before Nigeria was colonized by the British, social and political life was organized around dozens of kingdoms -including the kingdom of Borno, the Hausa kingdoms of Katsina, Kano, Zaria, and Gobir in northern and central Nigeria, and the Yoruba kingdoms of Ife, Oyo, and Ijebu in the southwest -and thousands of small communities. 4 Many of these were kingdoms administered by traditional rulers (based on family ancestry) with titles like Oba, Emir, Sarki, Shehu, Mai, Lamido, and so forth. 5 British colonialism (roughly 1861-1960) brought a system of indirect rule in which traditional and religious rulers -particularly in the northern parts of Nigeria -were allowed to retain their power and influence in return for cooperating with British political and economic interests (Blench et al. 2006: 11). Post-independence Nigeria saw the emergence of a 'mixed government' system, a symbiotic relationship between traditional and state-derived institutions (Sklar 2003: 3-25).
Today, the Nigerian government has a complex relationship with traditional rulers. Titles like the Ooni of Ife, Aareonokankafo of Yorubaland, Deji of Akure, Bobagunwa ilu Egba, and Alaafin of Oyo bring a great deal of respect, influence, and power in certain Nigerian communities. 6 According to Nigerian scholar Ali Yahaya (2005: 242), traditional authorities ' exercise considerable influence and are consulted by the federal, state, and local governments.' William Miles (1993: 31-50) describes how 'traditional rulers play useful roles in brokering between the people and the state, enhancing national identity, resolving minor conflicts, and providing an institutional safety-valve for often inadequate state bureaucracies.' And according to Blench et al. (2006: 24), 'The daily business of most traditional leaders is the settling of disputes involving family, communal, and religious life, and in many ways this is both useful to the community and valuable to the state, as many disputes that might otherwise clog up the court system are settled informally and usually in ways that accord with the local community's sense of appropriateness.' Meanwhile, another important group of non-governmental authority figures is seen among the leaders of Nigeria's Muslim communities. These are largely organized around two 'Brotherhoods' -the Qadiriyya and Tijaniyya. Although there are no accurate figures as to how many members each brotherhood may have, they are counted in the millions and can be found the length and breadth of Islamic West Africa. For this reason, their leaders can be significantly influential through their proclamations and religious edicts. They appear regularly on television and finance and run a range of religious and social programs through which they exercise considerable influence. Leaders in Nigeria's Muslim communities, like the Sultan of Sokoto, also exercise power through a system of Islamic courts in which disputes are mediated and punishments meted out to members of the community deemed guilty of a major offense. 7 To briefly sum up, the case of Nigeria illustrates how authority and perceived legitimacy to govern can be distributed between an official government and a system of locally-oriented cultural, religious, and ethnic allegiances.
From this perspective we therefore argue that in order to understand how illicit or private authority emerges, it is not sufficient to consider one non-state actor in isolation or in a dichotomy to the state. Instead, this article describes how and why we have to take into account the complex connections and interactions among different (violent) nonstate structures. To do so requires that we first overcome the widely-held assumption that only a government should be allowed to provide public goods, as this portrays any political authority by non-state actors as bad, threatening, and something that must be stopped. Our research suggests that some forms of complementary governance among non-state actors -even violent ones -can result in higher levels of security than a government can provide.
Moving beyond the dichotomized understanding of state versus non-state governance to governance that is constitutive of various non-state groups, the perspective put forward in this article is thus aimed to enrich the current debate on governance and security. After describing the various dynamic forms of these arrangements, our analysis concludes with implications for further research and policy.

Research on Political Authority among Non-state Actors
As noted earlier, recent decades have seen a tremendous proliferation of literature and scholarly debate on the nature of state weakness and its implications for political violence and human security. Researchers have sought to identify the most important ' conditions that give rise to and sustain environments that are susceptible to insurgency, terrorism, and other extremist violence and instability' (Thaler et al. 2013: xiii). Government legitimacy has often been cited as one of these conditions; according to renowned theorist Ted Robert Gurr (1970), as legitimacy of a political regime declines, their citizens are more likely to rebel (see also Habermas 1975). While a growing chorus of scholars has focused our attention on 'weak' or 'failing' states, it is the state-society relationship that is central to our understanding of political authority among nonstate actors, regardless of state weakness or strength. A government may have significant capabilities for the use of force in responding to security challenges, but when the legitimacy of those who deploy that capacity is undermined by endemic corruption among government agencies at all levels (including local police forces), citizens prefer to involve non-state actors in resolving disputes or addressing security problems.
Political stability and human security are thus seen as products of the extent to which the government is considered legitimate or effective. For example, a recent report on the Nigerian terrorist group Boko Haram describes how local citizens 'no longer believed in their country's political, economic, or legal institutions,' concluding that the 'most common and salient grievances [among those citizens] include corruption among political and economic elites, economic disparity, barriers to social and educational opportunity, energy poverty, environmental destruction, human insecurity, and injustice' (Forest 2012a). Similar grievances are expressed by citizens in other parts of the world where we find significant levels of political authority among non-state actors -examples include Pashtun warlords in the Afghanistan-Pakistan border regions, clan and tribal leaders in Somalia, Shiite militia groups in southern Lebanon, and (as examined later in this article) narcotrafficking groups in the Colombia-Ecuador border areas.
Economic conditions have also received considerable attention in the research on governance and non-state actors. Widespread poverty is frequently cited as an underlying reason for the emergence of a 'shadow economy,' loosely defined as economic activities that are unregulated and untaxed by a country's government, activities which are mainly underground, covert, or illegal (Naghshpour, St. Marie & Stanton Jr. 2007). They can include both monetary and nonmonetary asset exchanges, including the sale of goods and services, and normally require the support (or at least the acquiescence) of the local population, who in turn receives economic benefits. A shadow economy can address the fundamental need among a local population to make a living while providing numerous opportunities for non-state actors to expand their operational capabilities. Wherever illicit economic opportunities are more widely available than opportunities in the legal economy we can expect to find at least one type of non-state actor enjoying more power and authority than a state government. In fact, the collaboration or coexistence of various non-state actors is often the result of an arrangement of convenience to increase economic benefits for all stakeholders involved.
When analyzing political authority among VNSAs, the security implications of these political and economic conditions are fairly well understood. For example, VNSAs who facilitate these shadow economies are sometimes viewed locally as powerful, effective at getting things done, and as having significant resources. This translates for many locals into a belief that the VNSAs have good leadership and that they take care of their own and those who assist them -a belief that, in turn, helps the VNSAs attract new recruits and secure local support or at least acceptance of their illicit authority. This, by extension, also undermines the ability of governments to gather intelligence on the VNSAs because they are protected by a local population whose economic security could be jeopardized if the VNSAs were vanquished by the state authorities. Thus a shadow economy can provide an infrastructure for VNSAs to operate in, whereby its ability to meet its financial needs becomes easier and detecting or disrupting those finances becomes more difficult (Naghshpour, St. Marie & Stanton Jr. 2007).
In places with these kinds of enduring political and economic challenges, VNSAs are able to exploit 'gaps' in a state's governance. O'Donnell (2004: 41) explains how a democratic country can feature 'blue areas' which are characterized by effective institutions and governance, as well as 'brown areas' with ineffective governance and institutions: these 'brown areas' are subnational systems of power that have a territorial basis and an informal but quite effective legal system, yet they coexist with a regime that, at least at the national political center, is democratic O'Donnell (2004: 41). In brown areas citizens do not have 'full' citizenship but rather a low-intensity citizenship (O'Donnell 1993). Citizenship implies norms that foster the citizens' loyalty towards the state in exchange for state capacity and responsiveness. Citizens accept rules of appropriate behavior agreed upon by the state and themselves (March & Olsen 2008).
In zones of complementary governance with private illicit authority, these rules are defined by non-state actors and accepted by citizens. This entails 'shadow citizenship': 'a cluster of illegal institutionalized organizational structures that guide behavior' in territory governed by non-state actors (Idler 2012a: 97). Shadow citizenship differs significantly from what O'Donnell (1993) calls 'low-intensity citizenship,' where a local populace may have limited access to few (if any) effective government institutions, but can be empowered 'in terms consistent with democratic legality' to transform this into 'full, democratic, and liberal citizenship' (O'Donnell 1993(O'Donnell : 1361. In contrast, shadow citizenship is an entirely different sort of citizenship that is oriented towards non-state actors instead of the state. Thus, shadow citizenship not only compromises the validity of human rights and undermines democratic values, it also distorts a mutually reinforcing state-society relationship which ultimately is the basis of a state's legitimacy at home and abroad, and necessary for any state to be a credible partner in the international community's efforts to reduce global security threats. Embedded within this research on the intersection of governance and non-state actors are several assumptions. For example, it is widely inferred that non-state actors should not be allowed to be providers of security and governance. The conventional argument is that only a formal government should provide public goods and services, and by extension all governments must directly confront any non-state entity with perceived political authority. Naím (2010; characterizes non-state actors as entities that are inherently opposed to states. Thus, when discussing weak states, many scholars and policymakers have recommended various ways for the international community to embolden and empower legitimate governments to 'take back what is rightly theirs' (i.e. political authority in regions where nonstate actors have primacy). Scores of books, journal articles, and public speeches have articulated the need for strengthening the capacity of governments (through finances, weapons, and other means) so that they may supplant non-state actors and impose their official government authority within a particular contested terrain.
Another assumption in this body of literature is that when there are multiple forms of VNSAs within a specific territory competition among them is expected, perhaps inevitable. A closely-related topic in the security and conflict literature addresses how the presence of multiple VNSAs competing for power can have a negative impact on the local security environment (see e.g. Kalyvas 2006;Humphreys & Weinstein 2006;Balcells 2010; Hoover Green 2011). However, recent field research has found that non-state actors are not necessarily in competition against each other. Instead, complementary forms of governance and political authority among violent non-state actors also exist. Furthermore, in some instances complementary governance among VNSAs may provide higher levels of political stability and citizen security than would be possible by the existing government. This suggests a need to re-think common assumptions in studies of state weakness and governance.

Complementary Governance among Violent Non-State Actors
A recent study of political authority among non-state actors in the Andean region of South America illustrates the complex intersections of governance and non-state actors described above. For more than a year extensive fieldwork was conducted in the Colombian-Ecuadorian and Colombian-Venezuelan borderlands as well as the capital cities of Bogotá, Caracas, and Quito. These borderland regions are of particular interest because they feature relatively low state presence, ongoing internal armed conflict (Colombia), and the presence of criminal networks and other violent non-state structures (Ecuador and Venezuela) (Idler 2014;Idler 2012a).
Further, border regions are particularly prone to volatile non-state governance structures. This is because at least three characteristics make them attractive to VNSAs: first, their tendency for weak state governance systems; second, their low-risk/highopportunity environment arising out of their trans-nationality making illegal cross-border activities extremely profitable; and third, a proneness to impunity because they operate in spaces where two different jurisdictions and security systems meet (Idler 2014). These three characteristic features are particularly pronounced in Colombia's borderlands. First, the weak state governance systems in these regions are rooted in the Andean states' historical neglect of their borderlands. Due to Colombia's centralist political system, policies have largely ignored sparsely populated peripheries. In the neighboring countries political, economic, and social life has been concentrated in the centers as well (Restrepo 2009). Second, due to their geostrategic location, the borderlands are major sites of coca cultivation and processing and constitute crucial corridors of international drug trafficking routes (Bonilla & Moreano 2007). Hence, illegal cross-border activities bring huge economic benefits. Finally, their rough terrain and vastness make efficient control difficult, increasing the propensity for impunity in the area.
These circumstances must be considered against the backdrop of Colombia's intricate decades-old armed internal conflict in which many different stakeholders are involved. After more than a decade of violent struggle between Colombian Conservatives and Liberals, in the 1960s leftist guerrillas, most notably, Fuerzas Armadas Revolucionarias de Colombia -Ejército del Pueblo (FARC) and Ejército de Liberación Nacional (ELN), began to fight in an internal armed conflict against the Colombian government due to their discontent with social injustice (Leal Buitrago 2004). In the 1980s, paramilitary militias were formed to respond to the violence, but soon their activities proved to only further increase levels of violence in the country. From 2003 to 2006 the paramilitary umbrella organization 'United Self-Defense Forces of Colombia,' founded in 1997, was demobilized, leading to the emergence of smaller successor groups, paramilitary splinter groups and criminal groups labeled 'BACRIM' by the government (emerging criminal groups). Today virtually all these different groups seem to be involved in forms of transnational organized crime, including drug trafficking and arms smuggling, ventures which are particularly prosperous in border regions (Idler forthcoming in 2015).
Recent developments in the conflict dynamics driven by Bogotá's security policies further contribute to the significance of Colombia's border areas. In 2000 the US Congress approved a counter-insurgency and counter-drug initiative, the so-called Plan Colombia. It was developed by Colombian President Andrés Pastrana and subsequently formed part of President Álvaro Uribe's Democratic Security Policy (Chillier & Freeman 2005). While the Democratic Security Policy arguably made cities safer and weakened the FARC, it contributed to moving the conflict's impacts further out towards the periphery and beyond. Fumigations to eradicate coca partly resulted in the cultivations' shift beyond the borderline; the Colombian state forces' military operations against the guerrillas produced mass displacements across the country and became an additional reason for Colombian VNSAs to retreat to the borderlands where they could recover, reorganize, coordinate attacks, and engage in illegal economic activities. Among Colombia's borderlands, the Colombian-Venezuelan and Colombian-Ecuadorian borderlands in particular stand out due to the political and ideological cleavages between the three governments and their leaders. 8 Because of these cleavages, border cooperation deteriorated during the first decade of this century, increasing the impunity of groups operating in these borderlands (Idler forthcoming in 2015).

Data collection and analysis
From August 2011 through November 2012, anonymous, semi-structured interviews were conducted with participants ranging from academics, government officials, and members of the state forces to staff of the United Nations and non-governmental organizations, ex-combatants, community leaders (including afro and indigenous leaders), clerics, displaced persons, and refugees. Interview data were complemented by participatory observation in civil society events and local community gatherings, reviews of policy documents, media articles, and secondary literature, as well as the evaluation of surveys and databases.
The data gathered during this research project offered support for the characterizations of political and economic conditions that contribute to political authority among non-state actors. However, our analysis also revealed behavioral patterns among violent non-state actors in these regions that we call ' complementary governance.' Complementary governance arises from collaborative long-term relationships among VNSAs in which these actors assume governance functions. This has many benefits for the 'governing' VNSAs because they can carry out activities without being disturbed or sometimes even noticed by state actors. For example, terrorist groups can use the areas they govern as some form of safe haven in which to carry out terrorist training. Similarly, rebel groups can use these areas as zones of recovery and recreation and criminal networks can engage with relative impunity in drugs, arms, or human trafficking. Many of these illicit activities constitute economic ventures in which more than one VNSA is involved, and this becomes an important motivation for complementary governance among various VNSAs. Consider for example the cocaine business with its different supply chain phases ranging from the cultivation and processing of coca leaves and the supply of chemical precursors to national transportation and international trafficking. Different actors are involved in different stages and many of them overlap geographically.

Diverse forms of complementary governance
The patterns of activity described in this analysis are particularly exemplified by the groups and arrangements in the Colombian southern border department of Nariño and other areas along the Colombian-Ecuadorian and Colombian-Venezuelan border (Idler 2012b). There is a high concentration of VNSAs in Nariño for at least two reasons: first, developments in Colombia's security policies since the beginning of this millennium, and second, the department's geostrategic relevance. To begin with, the impacts of Plan Colombia were particularly drastic in Nariño's neighboring department Putumayo. Due to intense eradication efforts and the state forces' military operations, both coca cultivations and armed groups shifted eastward to Nariño where these groups have been gaining strength, particularly over the past five years. Furthermore, all stages of the cocaine production chain are present in Nariño. Favored by propitious climatic and geographic conditions, Nariño has the largest amount of coca cultivation areas of any Colombian department with 10,733 hectares of coca cultivation in December 2012, and this despite a drop from 17,231 hectares of coca one year earlier (UNODC 2013: 11). The territory is also heavily used for processing coca leaves into coca paste and crystallizing this paste. 330 cocaine base laboratories (cocinas) and 38 hydrochloride cocaine laboratories (cristalizadores) were destroyed in Nariño in 2012 (UNODC 2013: 86). There are further laboratories that have been detected but not destroyed. Finally, Nariño is also a hot spot for the latter stages of the cocaine business. It has both a land border (with Ecuador) and a maritime border (with the Pacific Ocean) and is thus home to two of the starting points of seven major international trafficking routes. In 2012, 16,816 kilograms of cocaine hydrochloride were seized in the department (UNODC 2013: 88).
At least three kinds of long-term relationships among the primary VNSAs exercising power and control in this border region appear to underwrite complementary governance: transactional supply chains, pacific coexistence, and strategic alliances. 9 Transactional supply chain relationships: In one prominent behavioral pattern, VNSAs participate in transactions along a supply chain, acting independently. The groups normally operate in a territorially segmented way and each group exercises economic, social, or political control in their respective territory. This modus operandi arises from the division of labor within the supply chain in which each group takes over specific functions. Enabling all parties to draw on their comparative advantages, this specialization helps increase profits from the different supply chain stages. Though the groups are hardly committed to each other, their financial or material transactions connect them in an indirect manner (Idler forthcoming in 2015).
The complementarity of governance is of a spatial nature, often in line with an urbanrural divide and based on the nurturing of a shadow economy. For example, in rural areas of Colombia's southern Nariño and Putumayo departments (along the border with Ecuador), the FARC's provision of public goods -including economic opportunities -directly contributes to their illicit authority. In Putumayo, the guerrillas established a presence several decades ago, created administratively and politically demarcated territories, and developed an economy based on coca cultivation. In these areas, communication and transport infrastructure is deficient or absent and many villages lack appropriate connections to nearby towns. Hence, rather than growing cacao or bananas which have to be transported to markets for commercialization, farmers prefer to grow coca which members of VNSAs purchase and pick up directly at their farm. Further, there are considerably more financial incentives in the illicit marketplace: for example, in 2009 a fisherman in San Lorenzo earned US$50 per week, but if he opted to work in the Colombian border zone as a raspachín (an expert in the collection of coca leaves), he earned between US$600 and US$800 per week (Idler 2012a: 100).
In urban areas along the Colombian-Ecuadorian border unemployment rates in the legal economy are high. The illegal economic opportunities offered by post-demobilized groups are thus sought-after. In such spaces, these post-demobilized groups can easily replace state functions, for example by offering work as messengers, informants, or drug mules. In such instances, the groups that control the cocaine business significantly affect regional economic dynamics by establishing and fuelling their own shadow economy. At the same time they need each other: the shadow economy of the cocaine business only flourishes if each stage of the cocaine business is doing well, which in turn requires governance capabilities of several actors. Thus, governance is complementary along the supply chain. 10 The various VNSAs involved in transactional relationships along the Colombian-Ecuadorian border provide two specific public goods that are particularly important for the legitimacy of an illicit authority: (perceived) justice and security. This illustrates how the group's control over the means of violence indeed supersedes the state's capacity. In communities in the Colombian municipality of Tumaco for example, the Autodefensas Nueva Generación, a group that apparently emerged after the attempted demobilization of Colombia's paramilitary groups, conducted 'social cleansing' by killing or threatening to kill ' antisocial' persons such as delinquents or drug addicts and thus impose security by their own means (SAT 2008). Another indicator of the illicit authority's control over the means of violence is that they restricted external institutions' access to the territory they control (Idler 2012a: 109). Similarly, in rural areas of Bajo Putumayo, delinquency was virtually absent by 2012 because, according to local residents, any delinquent would have been severely punished by the FARC. 11 As long as citizens stuck to the laws of the illicit authority it was relatively unlikely that they would experience physical violence. In that sense, some kind of 'shadow citizen security' might have existed (Idler 2012a). Brought about by violent, illegal means, however, it is really only an illusion or 'shadow' of citizen security, exemplified by the three youngsters killed and dismembered in 2010 by the Rastrojos gang because they had not paid extortion money (Defensoría del Pueblo de Colombia 2010). In rural communities along the Colombian-Ecuadorian border there have also been cases in which a VNSA told community leaders to only let ' outsiders' (such as humanitarian organizations) enter the territory if the ' outsider' did not address any human rights issues. 12 Pacific coexistence: In this arrangement, different VNSAs have tacit agreements of non-aggression and non-interference in each other's affairs (Idler 2012b: 78). They share territory, operating in a parallel manner without having any kind of interaction, and each VNSA exercises economic, social, and/ or political control within the environment they are embedded in. Thus, in contrast to a strategic alliance (where governance arises from the joint actions of two or more VNSAs), in the case of pacific coexistence each VNSA assumes separate governance functions (Idler forthcoming in 2015). Still, governance is complementary because, rather than competing with each other, they arrange themselves in a way which allows all actors to exert illicit authority by providing public goods, assuming control over the private use of violence, and by earning social recognition.
Situations of complementary governance arising from pacific coexistence can be found in the southern part of the Colombian-Venezuelan border region in the Venezuelan state of Apure. For example, the Venezuelan left-wing violent non-state group FBL, said to have been advised by the Colombian ELN, lives among the local population in Apure while participating in cocaine trafficking from Colombia to Venezuela. Colombian FARC and ELN are also present in this zone. Tacitly respecting each other's territorial limits, the three different groups control the political and economic life in their respective territories and ask the population to ' collaborate' financially with them. They are present both in rural areas and in cities such as Guasdualito, San Fernando, and El Nula where, according to the local population, they have established medical services, including surgical capacities, to care for injured people. 13 While on the one hand they provide public goods, on the other hand, in the eastern part of Apure and in the state of Amazonas, they control airstrips that are used to traffic drugs out of Colombia to the Caribbean coast, an important starting point of the trafficking routes to Europe and the US. 14 The money they derive from charging traffickers to use these airstrips is important for the VNSA's accumulation of economic power. However, it can also be used to facilitate the provision of services that help bolster their perceived legitimacy among the local population (Marti 2010).
In Apure, the ELN, FBL, and FARC have also been controlling the means of violence, superseding the Venezuelan state in this arena. The ELN for example is said to run a justice system for local communities that is applied to issues such as conflict between neighbors, thefts, and property damage. As a local cleric pointed out, the population of the village El Nula, for instance, turned to the VNSA in all kinds of conflict situations: 'If a cow gets lost, people run to the guerrillas. They always run to the guerrillas! …They intervene to provide even Solomonic solutions.' 15 Given that various VNSAs operate across the same territory, the relationship between them and the local community matters. People who have had a longer relationship with the ELN for example are more likely to turn to the group than to the FBL or FARC if they need judicial support. Therefore, social recognition plays an important role. Some people reported that the FBL receives less social recognition than FARC or ELN. Hence, people are less likely to ask them for justice. Nevertheless, each group has its 'niche' among a certain group of community members that allows them to exert illicit authority.
Strategic alliances: In a third form of complementary governance, VNSAs establish long-term commitments in which they share intelligence, revenues, and/or expenses. They operate in the same territory where they can (for instance) enter into 'war pacts' in order to jointly fight against a third group (Idler 2012b: 76). For the citizens that live in a territory where strategic alliances have been formed, governance functions are typically assumed by one unit: the allied VNSAs. In Nariño in 2011, ELN and FARC were said to operate jointly in the municipalities of Santa Cruz, Samaniego, Cumbitara, Las Llanadas, and Los Andes (among others) to combat state forces. They supposedly also had a strategic alliance in 2011 and 2012 in the area of Catatumbo. The two group names were painted on house walls as one single graffiti and would have been eliminated if one of the two groups had rejected them. 16 Also in 2011, two other VNSAs, the Águilas Negras and Rastrojos, apparently joined forces in Tumaco after having previously fought each other. 17 In these cases, complementary governance rests on the joint provision of public goods and the joint control of the means of violence as if it were one single actor.

Implications
As Hall & Biersteker (2002: 216) explain: The claim to authority of private illicit actors in the international system rests upon their capacity to provide public goods and their private control of the means of violence that competes with, or supersedes, the capacity of public authority. The social recognition of illicit authority is also essential to its emergence as private authority, not simply its possession of power.
Certainly, VNSAs in Colombia's borderlands have been jointly providing public goods and controlling the means of violence in transactional relationships, strategic alliances, and pacific coexistence. Though the complementarity in which these functions were carried out differed, in all cases the VNSAs achieved social recognition among the local population. In the case of transactional relationships, each actor achieved social recognition for the particular function it has assumed. In rural areas, for example, the FARC have contributed to building roads and health centers, which helps people tolerate negative aspects, such as extremely brutal punishment for non-compliance of the imposed rules. In urban areas, many citizens have appreciated the 'security system' which reduced assaults and robberies, even though the economic opportunities provided involved high risks (Idler 2014).
In pacific coexistence social recognition is arguably harder to achieve because the VNSAs operate in the same territory. Though not competing for governance they do have to compete for, or at least balance, the social recognition they are granted (Idler forthcoming in 2015). Thus, complementary governance in this form is much more fragile than in the case of transactional relationships and can easily transform into competition. Finally, in strategic alliances, citizens may perceive the VNSAs as one unit which is characterized locally as a legitimate 'governor.' This case is therefore most closely aligned to the rebel or paramilitary governance that has been examined by other scholars. However, close collaboration among VNSAs can also lead to the opposite result: if group identities are blurred and no clear new joint identity emerges, residents have difficulties in identifying a certain group and thus developing loyalties to it. As a consequence, social recognition becomes unachievable (Idler forthcoming in 2015).
Certainly, in all examples the VNSAs' social recognition is linked to the state institutions' absence or inefficiency. The provision of economic opportunities where the state has failed in this respect, for example, is one of the reasons why the FARC have gained the local population's support and are perceived to be legitimate in regions such as Catatumbo, situated in the north of the department of Norte de Santander on the border with Venezuela. Furthermore, some citizens declared that their community has been safer under the rule of the FARC than when the state forces were present because the FARC established order. Also, some people consider the post-demobilized groups' parallel justice systems more efficient: delinquents are punished immediately whereas the state justice system requires a longer process. Of course, to call it a true system of justice is debatable -in healthy democracies there are various means by which a population can constrain the public enforcers of law and order, while in the case of VNSA enforcers there are no clear ways for the powerless to prevent abuses by the powerful. Thus, social recognition and authority require the VNSAs to self-impose constraints on its use of force; those that do not risk losing legitimacy and, over time, the ability to ensure a compliant population.
In the case of insurgent groups, their social recognition also rests upon their ideological discourse. The FARC, for instance, emphasize that they fight for the cause of the exploited, for national liberation, and for socialism, referring to its origin as a Marxist-Leninist revolutionary guerrilla. Given the state's abandonment of many rural communities, it is not surprising that people here endorse the FARC and support their anti-state discourse. The FARC's historical political objective is connected to their internal behavior, a further factor that is likely to contribute to its social recognition. The profits that the FARC reap from the cocaine trade are mostly used to the organization's benefit and only to a limited extent enrich individual members. While participating in the drug industry certainly improves the lifestyle of some leaders, there are few cases of exorbitant personal enrichment. Disciplinary transgressions are severely punished and aggressions against the civilian population are only permissible when ordered by the leadership (Gutiérrez Sanín 2008: 14).
This behavior likely contributes to a local perception of legitimacy toward the guerrillas because it reinforces the perspective that their involvement in the drug industry serves the political goals they claim to fight for rather than personal greed. Further, it also demonstrates a kind of behavioral control within an organization that helps legitimate its monopoly over the use of force in ways that we may not see among government agencies. In essence, the ways in which diverse types of VNSAs are required to govern within their respective areas of influence have much in common. Legitimacy derived from these areas of activity allows VNSAs to 'govern' either on their own or in (tacit) agreement with others for whom stability and security are shared common interests. Significant swaths of territory can be governed in this manner in a complementary way, as evidenced in the Colombian-Ecuadorian and Colombian-Venezuelan borderlands.
In brief, the empirical evidence from the Colombian-Ecuadorian and Colombian-Venezuelan borderlands illustrate how longterm arrangements entail complementary governance based on illicit authority. These arrangements help provide some 'rules of behavior' to which the local population can adhere to avoid punishment. These circumstances can be described as shadow citizen security rooted in shadow citizenship (for a more detailed discussion see Idler 2012a). The most important point to emphasize here is that a shadow citizenry may find more security in areas with these kinds of long-term VNSA arrangements than in areas that are ostensibly state-controlled. Several recommendations for policy and further research can be derived from this analysis.

Conclusion
The overall intention of this analysis has been to fuel the debate on governance and security. Enhancing understanding of complementary governance among various VNSAs is a good starting point, we think, and indicates a long, but exciting path for future research. While the existing body of scholarly research would lead us to expect competition for power and control among non-state actors in these border regions, we instead discovered that in many cases they have established forms of complementary governance through complex, stable relationships with each other. Three such ' arrangements of convenience' stand out: transactional relationships, pacific coexistence, and strategic alliances. Operating under these arrangements, the VNSAs provide public goods, control the monopoly of violence, and are socially recognized by the local population. According to our analysis, the complementary nature of the governance functions takes shape in distinct ways that are related to each arrangement. Thus, taking into account the complexity and particularities of these relationships becomes crucial to understanding VNSA decision-making and behavior, particularly in terms of how VNSAs exert illicit authority through various forms of governance.
Security and political stability can be seen as being in the best interests of the VNSAs, thus explaining their relative commitment to providing public goods and services. Chaos and insecurity is clearly not in the best interests of either a government or nonstate actors, though for obviously different reasons. Arguably, as long as VNSAs have a vested interest in providing security and other public goods -and are enabled to do so by state weakness, indifference, or other factors -we will see examples of complementary governance for many years to come. With our research we move beyond the basic dichotomy of state versus non-state actors, and consider instead the more nuanced ways in which a variety of actors may contribute bits and pieces to an overall complex security environment. Perhaps it can be useful to envision a flexible continuum of governance relationships encompassing traditional forms of formal governance as well as shadow citizenry models described here.
Our analysis also points to the critical importance of economic opportunity in perceived legitimacy of a government. A state that is unable to provide robust economic opportunities for its people should expect that citizens will turn to a burgeoning shadow economy for their livelihood, and non-state actors can derive power and authority by supporting and protecting that shadow economy.

Policy implications
Confronting the security challenges resulting from VNSA social recognition will require more than simply a commitment to deploying a government's law enforcement or military forces to those areas. When focusing exclusively on the paradigm of competing governance, the 'state building' strand of thought promotes increasing state presence, fighting back the VNSA group by either military force or by winning back the hearts and minds of the local population (or by a nuanced combination of both). However, our analysis suggests that current state building approaches are ill-equipped to transform areas of shadow citizenship into regions with a citizenry engaged in a mutually reinforcing state-society relationship. States cannot simply replace non-state actors if they ignore the local non-state dynamics upon which the shadow citizenship is based. At the same time, approaches that suggest 'talking to the (one and only) enemy' miss an important dimension that has been revealed in this analysis: that in many areas -especially those which serve to engage in illicit economic activities that proffer financial benefits for many different stakeholders -more than one VNSA is exerting illicit authority. Only by understanding the complex relationships that these VNSAs have with each other can we identify entry points that help dismantling the complementarity of their governance, which, in turn, deprives VNSAs of their social base, rooted in their legitimization by the local population.
By depriving VNSAs of their 'shadow citizenry' the illicit actors' authority is reduced to crude power and it will become increasingly difficult for them to maintain a governance system that follows their own logic. Lacking the local communities' consent they will have to resort to violence to impose their rules. This however impedes maintaining a low profile; invisibility in the areas in which they are operating is no longer provided for. As a consequence, the territory may ultimately lose its attractiveness as safe haven for clandestine networks as the costs to maintain operations in these areas probably outweigh the economic benefits of their illicit activities. Meanwhile, drawing local participants away from the shadow economy and into the licit, formal economy must be part of a holistic strategy to undercut the base of support that enables the VNSA to operate. Without this critical base, the VNSA crumbles or, like the locust, will go elsewhere to feed off local resources.
Often governance not only is complementary between state and non-state actors, but also among an array of different VNSAs and other non-state structures. Therefore, it is insufficient to address one VNSA in isolation. A government's policies or actions against a specific VNSA in a region may benefit (or may threaten) other VNSAs, depending on the inter-VNSA relationships within that specific context. A government could create the conditions under which one (or more) VNSA would decide it is in their best interest to yield political authority to the government; conversely, a government's actions could lead a VNSA to determine that the best course of action would be to make things worse (from a human security perspective) by attacking various government or civilian targets. This is particularly the case in territories that have historically been under only nominal control of a state government. In sum, governments cannot simply replace the political authority of non-state actors in areas where there is more than one actor involved. Further, engaging with the most visible VNSA group is not sufficient to significantly change governance patterns. Instead, more innovative policies are needed that take into account the nature of the relationships among various VNSAs.
Identifying the specific form of arrangement -for example, transactional relationships, pacific coexistence, or strategic alliance -can lead to a better understanding about the nature of complementary governance. Is it segregated into different territorial units, unified into one joint 'governance actor,' or based on a fragile balance of social recognition of various groups? From this knowledge, effective responses can be developed. For example, our analysis suggests that a unified form of complementary governance may be the hardest to dismantle. Breaking the alliance among the VNSAs thus might be the first step to undermine their illicit authority. However, such an undertaking is likely to induce violence, thus policymakers would have to determine whether the costs are low enough for the perceived benefits of these actions. At a bare minimum, this research indicates that a government needs to acknowledge and recognize that VNSAs can earn political authority and social recognition among a local population. It is thus necessary for governments to understand the sources and nature of their illicit authority, how it is exerted, and who benefits and why. Then, a sophisticated understanding of these questions can help the government develop context-specific strategies to address the challenges posed by VNSA illicit authority.

Avenues for future research
There is a need to move beyond the dichotomized understanding of state versus nonstate governance, and to open up the debate to governance that is comprised of multiple VNSAs. What are the conditions that most likely lead to or enable complementary governance arrangements among VNSAs and non-violent non-state actors, and why? Pursuing research in this area could generate new insights into where else we might expect to find complementary non-state governance structures. Case studies from other regions around the world are needed in order to confirm, supplement, or contrast the findings of this analysis.
Further, VNSAs should be viewed as rational actors pursuing self-interests and making strategic decisions. While research is emerging on what influences those decisions (see e.g. Shapiro 2012;McCormick 2003;Forest 2012b;Martin & Perliger 2012;Felbab-Brown & Forest 2012) we have much more to learn. It must also be recognized that a variety of conditions and events can move a VNSA from one characterization to another: that is, on a given day Group A may be considered a competitor to Group B, and yet the next day these two groups are seen as collaborators. This may be due to decisions made by group leaders, in some instances reacting to events or changes in their surrounding environment. Any analysis that is based solely on static portrayals of a group and its political authority mistakenly overlooks the dynamic nature of socio-economic and political conditions that undergird that political authority. Further, a more nuanced understanding of VNSA behavior (both within individual groups and collectively among various groups, as seen in these examples of complementary governance) and decision-making leads to new insights about how VNSAs can be influenced.
The questions that need to be asked in future research include: How can we learn from the complex relationships, connections, and interactions among different VNSAs? What is the role of the state for these arrangements? How can we engage these multiple VNSAs for the purpose of moving from shadow citizenship to citizenship and from illicit authority to an authority based on the respect of human rights? More detailed research on the three categories of arrangements and on complementary governance are required in order to analyze these questions and to make firm policy statements.

Notes
1 Part of this research was assisted by the Drugs, Security and Democracy Fellowship Programme administered by the Social Science Research Council and the Universidad de Los Andes in cooperation with and funds provided by the Open Society Foundations and the International Development Research Centre, Ottawa, Canada. 2 Aware of the challenge of analytically distinguishing between 'state' and 'nonstate' violent actors, VNSAs are defined as a set of at least three individuals who are i) 'willing and capable to use violence for pursuing their objectives' (Schneckener 2006: 25); ii) directly or indirectly challenging the state's legitimate monopoly of violence by using or threatening to use violence illegally; and iii) 'shaped through an organizational relationship or structure that exists over a specific period of time' (Schneckener 2009: 8-9). 3 See also the U.S. National Security Strategy (2010) in which 'strengthening national capacity' is a central theme, and U.S. Department of Defense (2012: 1), which describes the need to ' establish control over ungoverned territories.' 4 For descriptions and photos of recent traditional and political leaders of Nigeria, please see the website: http://www.kingdomsofnigeria.com/ 5 For a detailed account of Nigeria's precolonial and colonial history, please see: Falola & Heaton (2008); Northrup (1978); and Djebah et al. (2003). 6 For example of community decision-makers, see Kuponiyi (2008: 239-243