Powerbrokers shape conflicts in Afghanistan, Mali, Libya, Iraq, Somalia and other places. As a consequence, all these countries have been destabilised despite massive international efforts. Millions of lives and trillions of dollars can be utilised without stabilising a political system, if the conflict is misunderstood. Focusing analysis on the outcomes of a system, such as corruption, smuggling, ethnic conflict, or lack of development can cause reactions by the most powerful actors within the system against the best intentions. This article aims to contribute to a theoretical understanding and discussion of conflict in weak states. More granular than one at the level of systemic phenomena, this analysis is focused on the actors, and the political structure in weak states. The article aims to improve on efforts to accurately describe these conflicts. After a theoretical introduction, this paper will apply a theory on powerbroker systems in Afghanistan, Lellnon and Mali. I will build the contours of the political system in each by looking at the very origins of socio-political life.

The current debate on civil wars and intra-state conflicts mostly focuses on greed vs. grievance (Levy and Thompson, 2010; Ballentine and Sherman, 2003) or motivation vs. opportunity, inequality (Stewart, 2008; Langer, 2008), culture (Langer and Brown, 2008) and organisation (Collier et al., 2003). Brown (2004) has tried to categorise the vast literature into four major categories of explanatory variables: structural, economic, political, and cultural factors. He concluded that none of these theories conclusively explains the observed behaviour and one should look to the actors, such as elites. Neorealism (and other classical IR theories), ostensibly intended to explain great power competition and behaviour between states, is rarely applied in conflict studies. The assumption is that life within a state is hierarchical but in the international system it is anarchical, thus different systems and different theories should explain outcomes. Structural realism assumes that the international system is populated by functionally equivalent unitary actors (states) and the structure of the system is only due to the relative distribution of power. This structure then socialises the behaviour of the actors. In weak and failed states, the system could also be described as anarchical and, thus, a structural realist theory could be applied if the powerbrokers fulfil the unitary actor assumption and become functionally equivalent units of the system. Waltz (1979) defined defensive neorealist thinking. He emphasised the socialising power of the structure of the (international) system towards its components. A system in which multiple powerbrokers exist in one state is analoguos to multipolarity and explains frequent alliance building and reshaping. This has been seen in civil wars by Christia (2012) and Otto (2018). Christia looks at civil wars in Afghanistan and Bosnia and concludes that alliances frequently reordered when the relative distribution of power changed. Her minimum-winning coalitions described an optimum distribution that fizzled out when alliances grew too large, as groups started to worry about their post-war rewards. Similar behaviour is highlighted in a continuum of politics through peace and war in the case studies below.

Sociology of Powerbroker Organisations

The Afghan tanzim is best known from the times of the anti-Soviet and US-financed Jihad. Since Rambo and Charlie Wilson’s War, they have been associated with rag-tag militias. However, the tanzims constitute the most important political parties in today’s Afghanistan. Hezbollah in Lebanon is similarly understood. The US sanctioned it as a military and terrorist proxy of Iran. The US believes the political wing only serves the military part (US Treasury, 2019). Several EU countries exclusively sanctioned Hezbollah’s military wing and hotly debate the nature of its political and social parts (Berti and Schweitzer, 2013). The tense debate between the US and EU shows that there is no unified view what to make of these organisations and how to assess their military, political and social components. The tanzim and Hezbollah are interpreted in their ethnic or sectarian terms, adding a social and ideological component. Neither of these actually describe what these powerbrokers are and how they act, but interpret the conflicts they are part of and throw an adjective down at them. By looking at the genesis and formation of these powerbrokers, I attempt to reverse this trend. Powerbrokers are rational and unitary actors within a weak or failed state (a self-help system). They are a social pyramid of patriarchic, pre-modern, kinship-based societal structures that build self-defence militias and band together during conflict with the centre or foreign powers. The powerbrokers foremost function is to defend the communities self-interest by brokering access to foreign rents or national resources. Powerbroker are the political-military organisations that represent the interests of one of these alliances of kinship-based social groups on the national level and towards foreign powers. Powerbrokers are proto-states within states that are mostly empty shells.

Where do powerbrokers grow? The answer is self-ruling communities, the actual state of nature for most of humanity today and all of humanity in the past. In Europe’s pre-modern past, the state intervened in limited cases, often only to levy for war or to tax. In either case, village, clan, or tribe were communally taxed. The absolute majority of people were farmers and bound to their place of birth. Most aspects of life, such as religion, trade, marriage, heritance, property, common goods and justice were administered internally by family heads that got together and selected village elders and organised self-protection. The feudal lords in medieval Europe at the outset of modernity were mostly busy keeping peace between self-ruling communities, securing trade routes and defence againt foreign forces. Social cohesion remained weak beyond the community level – just as in weak states today.

The outset of nationalism and capitalism in the 17th/18th centuries changed that for some societies. For Harari (2014, p. 434ff), this change is the displacement of the traditional family and community by the market and the state. These twin phenomena atomised society. They broke the community bonds that managed life for most people. Kinship commuities became citizens and consumers. By breaking down the communities and rearranging the individuls in much larger nation states, immense forces could be harvested. These brought the developing industrialist nations to other continents where their superior industrial workforces and armies colonised others. Acemoglu and Robinson (2012, p. 256ff) describe how African chiefs resisted the change in society that came with foreign technology, describing how ruling, traditional elites of non-industrialised societies wanted to maintain their traditional kinship ties as they ruled through these institutions. Most societies aceepted guns and railroads, but resisted social modernisation. Hence, industrialisation and socio-economic change is not guaranteed, but the modern nation-state was built on the industrialissed society, especially individualism, capitalism, bureacracy and militarisation.

This mismatch is the origin of the powerbroker – he is the link between tradition and the necessities of interacting with states in the world system. He builds a proto-state that promises not to change the social order from which it arises. Common religion, ethnicity, language and ideology increase trust and speed up cooperation. However, there are usually established ways of communication across those lines as well. It is practical to absorb smaller net-contributors as they cannot challenge any main group and seek protection. There are not only a few large social groups, but frequently dozens of smaller ones that seek protection (bandwagoning) and some larger ones might have small enclaves that need to make pragmatic choice1. Powerbrokers therefore often consist of a core social group (ethnic, regional or sectarian) but smaller outliers are readily accepted, because the organisation is defined to defend local interests, not primarily by e.g. ethnical considerations. Ultimately, power trumps ideology. Afghan powerbrokers have become masters at secret negotiations with disgruntled members of their rivals, and this will also be discussed in the Mali section in detail. This fractured bottom-up socio-political structure is what gives rise to powerbrokers. Understanding this fact can be vital. The Taliban victories in 1995 and 1997 were enabled by alliances – a Tajik powerbroker realigning led to the fall of the West and an Uzbek General delivered Masar-e Sharif from an Uzbek powerbroker to the Pashtun Taliban. The Taliban engaged in a deliberate pan-ethnic strategy from at least 2008 onward. Focussing on the ethnic aspect cannot explain these constellations, but analysing the powerbrokers as rational, unitary actors in an anarchic context can eplain this.

During conflict, the threat is higher and powerbrokers can convince their constituencies to increase size and scope. They ally, reach out to foreign nations for trade, arms, intelligence and training for their fighters. In peacetime, they are attracted to the national capital to compete with others for the resources that can be extracted by occupying high-level positions and maintaining foreign contacts.

After conflict, powerbrokers transform their administrative and intelligence apparatus into political parties. Militias can be integrated into police forces (United Tajik Opposition, Tajikistan), retained (Hezbollah, Lebanon), sent aborad (Tuareg, Mali/Libya), or integrated into ministries (Uzbekistan, Northern Alliance in Afghanistan). As long as the powerbroker does not attempt to reform his communities against their will, they support him. Patronage increases his status as it feeds hungry communities. Foreigners usually perceive this as corruption without seeing the systemic aspect, but securing relative power advantages for ones own group and alliance is the imperative of a self-help system. Corruption is a consequence of the structure of society.

While studying Afghan warlords, Giustozzi (2009, p. 108f) coded the village-level militia ‘vavassors’ and the regional commanders ‘vassals’ in reference to feudal systems. Warlords, equivalent to powerbrokers in this paper, are institutions that sit atop this three-level pyramid of local and regional self-defence forces born out of self-governing communities. These powerbrokers can be more personalised, as the Afghan examples show. In this case, a charismatic leader such as General Dostum or Gulbuddin Hekmatyar manages to maintain the leadership position, often for life. However, each rules with a council in which the vasall leaders sit, thus reflecting the constituent units. Others, such as Hezbollah, might have a charismatic leader, but a strongly institutionalised main body, and the Malian MNLA has a relatively weak leader, but strong council. Institutialisation largely depends on the level of foreign support and the difficulty of obtaining recruits, weapons, fuel, food, medicine, but also training, communications and intelligence. The more that flows through the broker, the better his position over time.

The fluid and pragmatic alliance behaviour does not lend to the theory of ancient hatreds, as Kaplan (1996) suggested in his study of conflict on the Balkans, but again, ethnicity or sect is an approximation by outsiders, and power -defined by the internal and external struggles- is the variable that explains competition and conflict in weak states. The threat for powerbrokers rises acutely if one group manoeuvers itself close to controlling all key positions and resources in a state. It could use this power to push the others to the margins and turn the system into a unipolar one. In the process, they tend to weaken, destroy, exile or kill other powerbrokers. Saddam Hussein and his Al-Bu Nasir tribe, the Al-Gaddafis in Libya, the Alawites united under the Kalbiyya tribe dominated by the AlAssad family of Syria, and the Al-Sauds in Saudi-Arabia are all examples of this.

Theory and Hypotheses

Powerbrokers render domestic politics anarchic. Where they are strong they contest the state. They maintain the means of power to defend themselves. When challenged, they are above the law. Powerbrokers means include (the threat of) protest, demonstrations, strikes, political violence, secession, even civil war. Their goal is to maintain defence forces without relying on taxing their own communities, thus drawing resources from the state, natural resources or sponsors. As each group tries to make the state pay for their own militia or dominate a strategic resource, competition ensues.

Competition under anarchy causes a security dilemma. I assume that powerbrokers are balanced on both sides, including the faction controlling the capital and centre of power, according to the theory of balance of threat (BoT). In this theory, states are balanced against the largest threat. I will adapt this to the proto states the powerbrokers build. Walt (1987) built this theory to refine Waltz’s (1979) balance of power with regards to alliance making by highlighting the components of power: aggregate power (population, wealth, industrial capacity), proximity (geography), offensive power and aggressiveness.

If any domestic ruler decides to centralise state authority it constitutes a threat to other powerbrokers (aggregate power). If the centre tries to expand the reach of a unified rule of law, it antagonises the self-governing communities at the root of all powerbrokers (aggressiveness). In these contexts, fighting corruption, levying taxes, nationalising industries and appointing generals all become inherently political as they change the relative distribution of power (aggregate power). Deploying security forces beyond the capital and appointing generals in provinces puts enemy forces at the doorstep of a rival (proximity). Demographic changes between social groups increase the threat (aggregate power), as does the introduction of arms by an external sponsor (offensive power).

My working hypothesis is that powerbrokers, built on self-governing communities, ally and bandwagon according to BoT. As domestic anarchy is an unusual concept I will test this working hypothesis by examining a derived hypotheses per case study and identify a) self-rule communities, b) the political-military nexus of powerbrokers within each system, and c) the related BoT alliance behaviour.

Case studies


Rural Afghanistan is known for its fierce resistance to outside influence. Famously, in 1842, after the East India Company deposed one ruler (Dost Mohammad Barakzai) and installed Shah Shujah Durrani, revolt soon occurred. King Amanullah Khan faced a tribal rebellion from 1919-1929 after trying to reform education, state outreach and change life in Afghanistan inspired by the European and Young Turk modernisation. The Afghan communists, the PDPA, decreed the emancipation of women and a land reform, touching upon tribal and village privileges and had to face a civil war from 1979-1989. Afghanistan has faced dozens of crises within the Palaces, with rulers often overthrowing their brothers and cousins, but when the Palaces tried to modernise outside the core cities, the communities organised tribal armies (lashkars) and marched on Kabul to end the infringement. When a more moderate, slow-reforming ruler has been installed, the country was able to return to peace in the villages and cabal in the Palace.

a) Self-rule communities: ‘Yaghistan’ survives

Less related to tragic events in history where tribal uprisings shaped great power politics, the behaviour of villages, clans and tribes as the genesis of socio-political life matters. Deobandi scholars in India wanted to incite a Muslim rebellion during World War I and move away from British colonial influence into the Pashtun areas of today’s Afghanistan and Pakistan. According to Haroon (2017, p. 147), they called this area Yaghistan, or land of the free. Haroon’s historical account of Pashtun tribes references how “there was no need for police or a judiciary as crime was dealt with immediately and justly by the village-level councils (panchayats), tribal social and political institutions.”

Two ways of modernisation, nationalism and political Islam, especially active during the 1940s and 1950s, were frustrated by the lack of enthusiasm that the Pashtun tribes expressed. Although the example that Haroon (2017, p. 154) gives refers to Pashtuns in what is now Pakistan, the social model and the resistance are comparable to Afghanistan: “Neither the Jami’iyyat al-‘Ulama-yi Hind nor Ghaffar Khan saw a way to introduce their objectives of individualism, the cornerstone of political modernity, in the Tribal Areas as they saw the underlying pattern of mullah-led politics as irreconcilable with a settled-district social order.”

We can identify the corresponding self-rule structures in the post-2001 era. The neo-feudal elites that represent the self-rule communities are described by Giustozzi and Noor Ullah (2006, p. 2) as “the village elders, such as arbabs and maliks – that regulate the sharing of water and act as intermediaries in the event of a conflict among villagers, have limited power and small clienteles, ranging between 5-70 men.” Furthermore, there are also khans, but that title is used for land-owning elites who form tribes that are more hierarchical, e.g. the Duranis that for centuries provided the Afghan kings. With regard to the extent of self-rule compared to obedience to the centre, it is difficult to make an estimate, but the National Solidarity Programme (NSP), initiated in 2003 by the Ministry of Rural Rehabilitation and Development and the World Bank, was supposed to link with the shuras and jirgas and Mullins (2009, p. 88) reports that “over 20,000 villages in all 34 provinces” have a form of local governance interacting with the NSP. Giustozzi and Noor Ullah (2006, p. 3) call the self-rule element “segmentary solidarity” and referencing Glatzer state:

“[p]olitical leaders can hardly build their power on the tribal structure alone since that is an egalitarian one. They continuously need to convince their followers and rivals of their superior personal qualities, and must procure and redistribute resources from outside the tribal realm. Their followers expect material or symbolic advantages from them, and in times of political chaos people demand that their leaders provide security. Clients may quickly be disappointed in a khan or commander and may switch overnight to a different one.”

They argue that where tribalism is unchallenged and little outside interference forces behavioural change, either induced through resource competition or by a lack of outside threats, tribalism even inhibits the creation of the powerbroker proto states. In summary, we can identify a continuity of self-governance – Yaghistan is alive and dominant among 75% of rural people.

Self-rule communities are usually hardly identified as such, because they talk about their kin groups, ethnicity or religion and less about the conflict with others. What can be assessed by legal scholars under the term legal pluralism or alternative dispute resolution is the outside perception of these systems and provides operationable features. A 2004 report by the International Legal Foundation (ILF) categorises different localised systems of traditional justice and assesses the following regions: The South and East, Hazarajat, Nuristan, and the Northern Region, where it distinguishes between practices in Badakhshan, Takhar, Kunduz, Samangan and the three provinces of Balkh, Jawzjan and Sar-e Pul. This report does not mention the West (Herat), but already entails the majority of ethnic, linguistic, sectarian and regional groups in Afghanistan. The institutions in question, jirgas and shuras, decide on family law (e.g. inheritance or marriage) and the village’s common goods, such as water and common land. They also rule on crimes against persons as well as properties. They exclude state law explicitly. Pashtun norms under the Pashtunwali (way of the Pashtuns), in particular, aim for the solidarity of the group by asking for forgiveness. This act, known as Nanawati (ILF 2004, p. 10), is at the centre and aims to restore the status quo ante before the wrongdoing and maintain the cohesion within the village or clan.

USIP (2010) has reported on the prevalence of self-rule communities in Afghanistan. The authors actually observe BoT behaviour even at this level:

“More alarmingly, work by the Afghan government and international community has disrupted and distorted local political processes. Military negotiations with tribal elders have in some places incited violence between local groups, particularly when groups feel as if funds and power are not being distributed equitably.”

These conflicts have baffled international observers for decades. They can be well explained if we consider each village or clan as a functionally equal socio-political unit and the shura or jirga as its representative political body and local forms of law such as the Pashtunwali2 as the observable way to organise self-rule. These local political units are in constant competition with neighbouring villages, clans or tribes. If these competitors receive support from either the government or international donors, or one powerbroker, the competing locals will likely feel the need to join a faction themselves, probably a rival.

b) Political-military nexus

The historical communities observed by Haroon (2017, p.148) in the last section were known to create tribal armies, so called lashkars. It “was argued that it was only through such a compact between the ethnically connected segmentary groups that the structural and governmental independence of all could be achieved.” The word lashkar seems anachronistic today, but the post-2001 incorporations are familiar as arbakis or mahaz.

Arbakis (sometimes Arbakai or Arbakay) are village and clan level armed militias, tasked to enforce the rulings of jirgas. The term might describe the warrior bands of Aryan tribes in wartimes (ILF 2004, p. 11). They are also the source of self-rule as they guarantee order internally and external protection. During the 1979-1989 war against the Soviets and the ensuing civil war in the 1990s and with the intense threat of a determined modernising, communist government, arbakis banded into larger militias under charismatic commanders – the vavasor or sometimes a vassal in Giustozzi’s neo-feudal model.

Post-2001 arbakis have been included in several schemes to man local units under the control of the Afghan state. The Afghan Local Police (ALP) has been severely criticised for failing to achieve security and stability as “[t]he creation of the ALP under the influence of local commanders, influential figures, and foreign forces without the supervision of the government can make the ALP members disobey the government and legal institutions of the central government.” (AIHRC, 2012 p. 5). Despite the uniforms, the ALPs were still tribal and loyal to their jirgas or local commanders. They police their ‘own’ people well and abuse their rivals.

Mahaz literally means front. The term was used by some of the Islamists in the 1980s. It is still regularly used by the Taliban and Jihadi groups. Giustozzi’s (2019, p. 267) identified Taliban chain of command can be understood as five regional Shuras under one central leadership. The Mashad and Northern Shura, for example, organise three Mahaz each with a district and village units as a parallel system to hold territory. Tomsen (2011, p. 218) describes how Massoud’s Northern forces were divided on three levels, “village militias trained to fight locally; well-armed light infantry units of about eighty fighters each, capable of moving rapidly within the Panjshir and to adjacent areas; and a ‘central’ force of heavily armed thirty- to forty-man commando units.” Even though ideological enemies, the Northern Alliance and the Taliban have similar socio-political origins and equal military organisation.

The political-military bodies that rose during the anti-Soviet struggle are usually named tanzim. As the 1978 unrest turned into armed revolt against the communist’s intrusion into village life, the revolting local arbakis organised themselves into regional mahaz. Afghanistan and Pakistan had used Pashtun communities on both sides of the border to harass each other. In the 1950s and 1960s, it was mainly Afghanistan that employed these tactics (Tomsen, 2011, p. 242f). By 1975, Pakistan had turned the table and organised Islamists that fled Kabul university.

Pakistan channelled US and Saudi money via its own intelligence service, the ISI, to Afghan groups. “The brigadier [in charge of ISI’s Afghan bureau at the time] revealed that 67 to 73 percent of all CIA- and Saudi-funded weapons were delivered by the ISI to the main Afghan extremist party leaders whom Zia favoured: Gulbuddin Hekmatyar, Burhanuddin Rabbani, Yunus Khalis, and Abdul Rasoul Sayyaf.” (Tomsen, 2011, p. 242f) Connections were made between refugees that established networks in Pakistan (and Iran for Shias) and the Pakistani religious-political Jamaat-I Islami and the Jamiat-I Ulema-I Islam. These groups allowed Saudi religious organisations to operate near the Afghan border, such as the Muslim World League (MWL) office in Peshawar which was led by Osama bin Laden’s brother-in-law in the 1980s. The Afghans had to organise liaisons with the ISI, bring recruits to the training camps, engage with wealthy Golf donors and Muslim Brotherhood infiltrating the MWL, maintain media and PR offices as well extensive logistics and supply lines. The political layer organising foreign support, intelligence and administration on top of small regional armies came to be known as tanzim. The lines of influence go as follows:

Village/Clan → Arbaki → Mahaz ←→ Tanzim ←→ Foreign world/Capital

In parallel, the communist regime established similar institutions. As Giustozzi (2009, p. 53f) writes, a jirga in Kabul in 1983 established the “Regional Force”. The tribal representatives in the jirga told the government they did not want to join the regular forces. Again, self-rule logic prevailed. Local notables had no choice but to seek assistance when some of their rivals joined the fledging Mujahedeen tanzim but haggled to maintain their self-rule and convinced a desperate government to arm them. The groups supported by Pakistan were named the Peshawar 7. They, their splinters and former opponents of Iranian-supported tanzims, and the militias founded by the communist government, still make up the body of powerbrokers in Afghanistan. Socialised by the conflict they are indistinguishable at the regional level and have similar organisations: political party, militia, social foundations, education and media while governing and administrating their strongholds. They are functionally equal proto states.

a) Alliance behaviour

Current US Special Representative, Ambassador Khalilzad (1997, p. 37), has described the situation in Afghanistan as “anarchy”. Afghanistan’s shifting alliances (Tomsen 2011, pp. 3–22, 619–652; Lee, 2018, pp. 658–661) have rarely been studied with few notable exceptions (Giustozzi, 2009; Christia, 2012).

The alliances of the 2014-2019 National Unity Government framed by presidential elections highlight the power struggle. The table below documents the changing nature of the alliances.3 Powerbrokers without formalised parties, such as the Barakzai and Achakzai tribes in Kandahar, are simplified as ‘tribal South’. Some alliances changed throughout the year and the columns are an approximation. Some parties, such as Jamiat have more than one significant powerbroker and can be split but are approximated with their majorities. Mehwar-e Mardoum was officially established in 2017 (Afghan Analyst Network, 2017b), but the powerbroker organisation represented existed before. Hizb-e Islami emerges in 2017 after the peace deal with the Afghan government. This table excludes the groups that are in armed conflict and oppose the current political system.

Table 1

Alliance Structures of Afghan Powerbrokers 2013-2019)

YearAlly governmentOpposition IOpposition IIOpposition IIIOpposition IV
Tribal South
Parts of Jamiat
2018Hizb-e IslamiJamiat
Tribal South
Parts of Jamiat
Parts of Jamiat
Table 1b

Actors and their characteristics referred to in Table 1 (assessed dominant ideological background of powerbroker in bold

IconNameEthnic orientationDominant regionIdeology
NIFANational Islamic Front of AfghanistanPashtunEast, KabulSufi, Nationalist
ANLFAfghani National Liberation FrontPashtunKabul, SouthwestSufi, Nationalist
JamiatJamiat-e IslamiTajikNorth, KabulSunni Islamist
JumbeshJumbesh-e MilliUzbekNorthwestLeftist, Nationalist
MehwarMehwar-e MardoumPashtun, DurraniKabul, SouthTraditional-Royalist
WahdatWahdat-e MardoumHazaraHazarajat, Kabul, NorthShia Islamist
IttehadIttehad-e IslamiPashtun, GhilzaiCentre, Kabul, KunduzSunni Islamist
Tribal SouthBarakzai, AchakzaiPashtun, DurraniSouthTraditional-Royalist
Hizb-e IslamiHizb-e IslamiPashtun, GhilzaiCentre, East, NorthSunni Islamist

The election years (2014, 2019) tend to form bipolar systems, due to the nature of the contest focussed on a winner-takes- all electoral system. As the winner gets the Presidency and the potential power to control the armed forces and police and appoint all the ministers and provincial as well as district governors, the polarisation at this time is also consistent with BoT-induced balancing behaviour. Moreover, there is no observable correlation between any particular regional, ethnic or ideological orientation and the actual alliance structures formed by the powerbrokers.


Lebanon is usually described in sectarian terms, in the news (Al Jazeera, 2019), by think tanks (Henley, 2016; Yahya, 2017), in academic literature (Fakhoury, 2014; Salloukh, 2016; Lijphart, 1977; Horowitz, 1985), even by its own constitution. Ghosn and Parkinson (2019) and Majed (2016) argue this could be due to the repetition of the same methods with the effect of reproducing the same narrative. Similar to Afghanistan, a look beyond the surface of the national powerbrokers and towards their constituent components reveals a very localised socio-political environment at play.

a) Self-rule communities

Gulick (1953, p. 367) sees villages as the “discrete, compact social unit whose inhabitants constitute a strong in-group vis-à-vis all outsiders.” He sees “kinship, religion, and devotion to the land” as pillars of society. Here, patrilineal kinship is dominant for the “basic loyalty group called into play in the family feuds famous in the region” (Gulick, 1953, p. 368).

The Ottoman millet system4 and French intervention created modern sectarianism of Lebanon. The millet system turned feudal elites from landowners to sectarian community leaders. Calfat (2018) reports that:

“[r]eigning over an empire that was incredibly diverse ethnically and religiously, the Ottomans relied on local feudal elites and community-entrenched dynasties to rule it. These political leaders, known as zu’ama, used their status to provide protection and patronage (wasta). In fact, some of these same dynasties constitute the same sectarian elites in present-day Lebanon and are part of the ruling elite coalition.”

Some researchers categorise the zu’am as feudal landlords and the clans as similar but distinctly different organisations, based on either patronage or blood relationships. Wege (2010) makes that distinction between the feudal Southern Lebanese Shiites and the more clan based Bekaa valley. The two systems are similar and produce the same outcome. The patron, “the zaim, generally gets elected in exchange for his ability to provide services to his community” (Lefèvre 2014, p. 7).

Majed (2016, 2017) describes how the more than 15 (sectarian) communities of Lebanon all have their own parallel customs and laws, based on sectarian ideas. “Therefore, within this system, important aspects of one’s social life (e.g., marriage, divorce, custody, birth, death and inheritance) are bound by the laws of religious courts; and most routes to political representation are tied to one’s sectarian identity (given the sectarian quota system).”

Inside each sectarian group, clan-rivalries exist. The Druze, for example, consisting only of 7% of the Lebanese population have two factions: the Jumblatt and Yazbaki families. Whereas Kamal Jumblatt was a national leader, his son Walid is seen as a sectarian champion (AbuKhalil, 1985, p. 31). Both are “lord of the Mukhtarah” (AbuKhalil, 1985, p. 32) as the heads of the Jumblatt family are known by their party. The party is called progressive and socialist, while the leader is a hereditary feudal lord and the reforms in the party programme are meant for others not their own leadership. Currently unable to challenge the Jumblatts, the Yazbaki have joined the Syrian Socialist Nationalist Party (AbuKhalil, 1985, p. 33). Internal rivalry is in this case more important than Druze solidarity, often said to be extremely strong due to the secretive religion and centuries of surviving persecution. There are, therefore, more powerbrokers, meaning more actors able to affect national politics through their political and military institutions, than sectarian groups, as the larger sectarian groups are not necessarily united under one umbrella organisation. Even within the same sectarian group, power defines politics and conflict more than ideology or religion.

There are incidents where individual clans have used their own forces for narrow self-interest, feuds, and extraction of resources. The Meqdad clan in Beirut and the Bekaa kidnapped Syrians and Turkish citizens in August 2012 to put pressure on Turkey and the Free Syrian Army (LBC, 2012). The following quotes are from clan members in an Al Jazeera (2012) article:

Hussein “Abu Ali” Meqdad, a member of the Meqdad family, told Al Jazeera the armed wing of the clan consists of “1,500 bodies, and another 1,000 on stand-by.” “In Lebanon, every family needs to protect itself while there is no government,” he said (…) Author Hamadeh said that the clans “are the least attached to political parties, including Hezbollah and Amal. (…) They abide by certain rules that they themselves have created, and have their own judge to look over their feuds,” he said. “The state does not get involved.”

Media reports claimed the Al-Jaafar clan had established a military wing (Naharnet, 2012), even though they initially denied having such a wing, and the Jafaar clan is accused of destroying army checkpoints in later incidents (DailyStar, 2018). Even before the 2011 Syrian civil war started, the Jafaar clan was accused of challenging the military on their home turf (LA Times, 2009).

Already at the level of clans, the impunity sets the stage for anarchy as the following example illustrates. “Baalbek-Hermel, which lies near the Syrian border in Lebanon’s Bekaa Valley, is widely under the influence of clans. According to the area’s governor Bashir Khodr, there are 37,000 arrest warrants against 1,200 wanted individuals” (Asharq Al-Awsat, 2018).

The same behaviour occurs among Sunni groups. A Carnegie report by Lefèvre (2014) found clan politics at the root of violent crises in Northern Lebanon (Tripoli). “And three large clans, the Mori, Nashar, and Aswad families, wielded influence in the popular area of Zahriyeh, the old city’s souks, and the Sunni stronghold of Bab al-Tabbaneh, respectively.”

b) Political-military nexus

Kin-based clans and neo-feudal landowners (zu’ama) are the origin of social power in Lebanon. Most maintain small militias, but Lebanon is said to have functioning armed forces (LAF) and only Hezbollah as the exemption to the disarmament of forces after the civil war. That gives three vectors to explore. Firstly, whether the LAF and other security forces are in fact shielded from clan-based politics, and secondly, whether Hezbollah’s forces are affected by clan-based politics, and thirdly, whether there are other clan-based security organs.

The LAF is seen as reborn after the civil war and, especially after the Syrian withdrawal, as the only national institution. This seems to be confirmed by surveys of the population (Geha, 2015). Nerguizian (2015) notes a phase of intense client competition between the two political alliances (March 8 and March 14, further below) with each trying to control the LAF but blocking the other. The explanation probably lies in the long-time control and shielding that the Syrian military and intelligence had over the LAF. Veen (2015, p. 16f) states that:

“[t]he legacy of 15 years of Syrian rule perpetuates this state of affairs because Syria basically used the Lebanese security forces as a gendarmerie-type extension of its own capabilities: good enough but not too good. In addition, officers originally appointed during Syrian rule still represent a significant section of the highest ranks of the LAF, which inhibits reform.”

Veen concludes that LAF intelligence is collaborating with Hezbollah. Furthermore, the General Security (GS), tasked with border security and intelligence, is under Syrian and Hezbollah influence. The Internal Security Forces (ISF) are influenced by Sunni groups, i.e. the Future Movement of the Hariri clan. Veen (2015, p. 21f) clearly sees his interviewees saying that other than sectarian quotas at entry-level, the promotions to more senior levels of the armed forces are clearly patronage based.

Powerbrokers deliberately keep the LAF underdeveloped so it cannot challenge their interests. The situation is different from Afghanistan, where powerbrokers prefer to seek access to the patronage opportunities within the armed forces to let the state pay their own militias. In Lebanon, the core of the LAF seems exempt as it is kept weak, but security and intelligence are under the influence of powerbrokers.

Eisenstadt and Bianchi (2017) point out that Hezbollah’s structure is heavily influenced by clans.

“However, our new examination of the biographies of more than 2,100 Hizballah fighters “martyred” in combat from 1982 to July 2017 highlights an additional, key factor. Hizballah has relied on family, clan, and local solidarities (or asabiyya, the concept of in-group solidarity made famous by the medieval Arab thinker Ibn Khaldun) to recruit members and build motivated, cohesive, and effective units.”

Similarities abound beyond the sectarian divide. Iran’s role in the creation of Hezbollah is well-researched but the fact that “the Pasdaran initially sought out the Shi’a clans of the Beeka, rather than south Lebanon’s Shi’a, because for various cultural reasons the group had been excluded from the upper ranks of the AMAL movement” (Wege, 2008) is less known. Wege continues to explain that “[t]he Musawi and Hamadi clans became the core clans of the embryonic Hizballah organization, and this structure magnified the difficulty for hostile services attempting to penetrate Hizballah. It made the exercise akin to penetrating a family. “Eisenstadt and Bianchi had identified Musawi as the most prominent last name among the fallen Hezbollah fighters until today. Other clans include the “Hamiya, (…) Aqeel, Shahadehs, and Ezzedeens” (Wege, 2010, p. 30).

The clans are grouped into three regions, Bekaa valley, South Lebanon and Southern Beirut, each with different local elites. That is the socio-political root of the elaborate powerbroker organization leading to Hassan Nasrallah. A socio-political structure very similar to Afghanistan emerges:

Clan/Zu’ama → Militia → Militia cluster ←→ Powerbroker ←→ Foreign world/Capital

c) Alliance behaviour

Lebanese powerbrokers, e.g. Amal, Hezbollah, Phalangites and the Future Movement, have their origins in the clan-based sociology of Lebanon, its self-defence militias and the need to band together during the civil war and the Syrian and Israeli occupations. Two generations later, the political class that has become accustomed to leading the party wings of the powerbroker organisations are passing their privileges on to the next generation. Before the May 2020 elections, the Arab News (2018) published a report stating that “[n]early a quarter of the 128 seats are expected to be passed on from an older relative to another member of the family, as the country’s politics of clans and dynasties shows little sign of fading. Of these, 19 candidates are standing for seats currently held by a father or mother.” The report specifically mentions the Jumblatt (Druze), Mouawad (Maronite), Frangieh (Maronite) and Tueni (Greek-Orthodox) clans. The Frangiehs are in the fourth generation.

A small country full of powerbrokers as armed, functionally equal proto states induces the anarchy that incentivises BoT-consistent alliance behaviour. The anti-ideological aspect of Lebanese alliance-making is seen in the two alliances that have dominated the political landscape since 2006: the March 8 and March 14 alliance.5 The table shows no ideological or sectarian clusters. Both alliances contain leftist, liberal and conservative groups of each sectarian branch. The BoT defines alliances. The largest Maronite faction, the Free Patriotic Movement of President Aoun is in an alliance with Shia Hezbollah while smaller Christian groups ally with the Sunni Future Movement. The exception here are Amal and Hezbollah, the two largest Shia groups. They are temporarily in one alliance, but their bilateral history is full of conflict and tensions (Mroue, 2020) mounted when Amal refused to send troops to Syria (The Arab Weekly, 2016).

Table 2

Lebanese national political alliances after 2018 (data taken from Pabst, 2015), Mizrahi (2019), Shavit (2018), Orion (2017).

March 8March 14
Free Patriotic MovementNationalistMaroniteFuture MovementConservativeSunni
AmalConservativeShiaLebanese ForcesNationalistMaronite
Syrian Social Nationalist PartySyrianLeftist, some DruzeIndependence MovementNationalistMaronite
MaradaSocialismArmenianNational BlocLiberalMaronite
Armenian Revolutionary FederationNationalistArmenianNational Liberal PartyLiberalChristian
Lebanese Democratic PartyConservativeDruzeDemocratic LeftLeftist
Ba’ath PartyBa’athistHunchakianSocialismArmenian
Arab Democratic PartyBa’athistIslamic GroupIslamicSunni
Union PartyNasseristFree Shia MovementIslamicShia
Arab Liberation PartyAnti-ZionistSyriac UnionAssyrianAssyrian
Toiler’s LeagueSocialistShurayaAssyrianAssyrian
People’s MovementLeft

The Cedar revolution empowered the Future Movement. As the Syrian war brought Sunni refugees into the country, the Maronites of President Aoun formed an alliance with Hezbollah as they felt threatened by the demographic shift. As this situation is ongoing, the alliance has already lasted several years.

The Lebanese political system is based on clans that self-govern and defend their interests. Armed clans as building blocks of powerbrokers were identified within all sectarian groups (Maronite, Sunni, Shia, and Druze). Depending on the level of external threat, they can band together to form larger militias and parties, but into more than one alliance in the case of clans of the larger sectarian groups. None of the alliances have to be pure in an ideological or sectarian sense and they are frequently reshaped depending on the BoT between the powerbrokers against any assumed political, ideological or sectarian preference. The LAF’s core is an exception. It maintains a national outlook as it has been influenced by outside forces and is deliberately kept at capacity, where it can be managed by the political system. Other security and intelligence services follow the pattern of powerbroker influence and patronage.


Mali has been in disarray since 2011, but the roots of the conflict are older and can again be found in the socio-political system. There is a civilisational division line running through Mali where the social groups in the South are more African and the North is more Arab and Northern African. I will focus on Northern Mali as a subsystem and focus on the 2012 Tuareg uprising that led to international intervention.

a) Self-rule communities

The Tuareg of Northern Mali are at the core of the National Movement for the Liberation of Azawad (MNLA) and the Islamist groups in the Sahel. There are smaller ethnic groups along the river Niger, such as the Songhay and Peul, and all interact with the Malian state. The conflict is either understood as colonial (the French border denied Tuareg their state), ethnic (Tuareg and Arabs vs. others), or religious (Islamist Tuaregs fight for Sharia against the liberal South). Looking at the components of each powerbroker, BoT is a better explanation.

The Tuareg are a tribal people that live on the edge of the Sahara on either side and are known to be traders that traverse the Sahara. Tuareg are divided into confederations consisting of noble and commoner clans. Even among the higher clans, some are more revered. The noble clans are powerbrokers of multiple lower clans which act as vassals. An adaptation of Pezard and Shurkin (2013, p. 4) illustrates clan hierarchies. The purple shaded fields indicate elite clans known as Ifoghas.

Table 3

Tuareg clan structure

Noble clan →Commoner clan
Kel Afella, Iriyaken, Imezzekaren, Kel Abalaq, Telindjest, IwindchedjenIredjanaten, Ibakayaten, Imakalkalen, Kel Talgatghat, Chemenammas, Chebel, Idabeylalen, Kel Ghala, Kel Edjerer, Kel Tidjereren, Ibeyenbaran, Igdalen, Izzegaghen, Iboghelliten
Ifergoumessen, Kel OuzzeynIdarabaten, Chebel win Adjous, Iwinchedjen, Imakalkalen
Kel Taghlit, Kel EssoukIgdalen, Idabaggaten, Imakalkalen, Ifarkasen, Tedjer, Aselloum, Iboghelliten
Taghat Mellet, Kel Oukenek, Kel TelabitIbelbetiyen, Dag Inahaket, Tachrerat, Dawsahak, Achakatlem
Idnan, Win Talkast, Win Taïtoq, Kel Ahelwat, Kel Terghrecht, Inheren, Dag AselleselImghad win Elewidj, Imghad win Adjous, Ighanneghasen
IbatanatenKel Takhakayt, Kel Terghecht

Pezard and Shurkin (2013, p. 3) show that from among the Noble clans a chief, amenokal, is elected and that “amenokals have derived their legitimacy historically from their warrior status-they protected vassals.” Following a similar socio-political structure, the Arab tribes of Northern Mali are divided into two major confederations, the Berabiche and the Kuntas. They consist of leading tribes and vassals. The Songhay consist of villages that elect their own chiefs. The Peul are organised similar to the Tuareg.

Tuareg nobles have always carried arms. Their monopoly on arms and a loyal warrior caste (imusay or imohar) within clans is how they maintained dominance over commoner clans and lower castes. Arms and independence are deeply rooted in Tuareg clan identity and the basis of society’s stratification. The Tuareg possess famous traditional armaments such as the takoba (sword), daggers, lances and shields, but fell behind when firearms were introduced. That is why the return of trained fighters and weapons in 2011 was important for the 2012 rebellion.

b) Political-military nexus

Mali’s North had four major rebellions (1963-64, 1990-96, 2006-09, 2012-13). Each was led by a powerbroker. The 1963 rebellion started when Mali tried to alter the socio-political structure of the Tuareg. Comparable with the Afghan communists, Mali tried to modernise tribal and rural peoples and sparked resistance. Mali tried to settle the clans, ending their nomadic lifestyle and change its hierarchies. The rebellion was led by noble clans (Kel Adagh) that wanted to stay in power. While the amenokel was passive, his brother led the rebellion (Pezard and Shurkin, 2015, p. 11). Mali won and executed many rebel leaders while others fled. The fleeing clan leaders built more sophisticated political, military, diplomatic organisations in Algerian and Libyan exile and led subsequent uprisings (Chauzal and van Damme, 2015, p. 31). Gaddafi integrated clans into his African Legion. Some of them returned in the late 1980s, and most after Gaddafi’s death in 2011.

The 1991, 1992, and 1995, peace agreements were signed by the Popular Liberation Front of Azawad (PLA, Tuareg), the Popular Movement of Azawad (MPLA, later MPA, Tuareg), Ganda Koy (Songhai), the Arabic Islamic Front of Azawad (Arabs), and Mali. The 2012-2013 groups are similar in name and identical in nature. Today’s National Movement for the Liberation of Azawad (MNLA) and Ansar Dine are led by the same clans as were the PLA and MPLA/MPA. The Ganda Koy did not even change their name. Fighting did not stop, despite peace agreements with Mali. In the 1990s, the Malian government created Kidal as a province to appease its Northern rebels. Soon after, the Kel Adagh fought other Tuareg that wanted to challenge the Kel Adagh primacy together with Arab and Songhai groups.

The Ifoghas are currently split in two powerbrokers, the MNLA and Ansar Dine. The MNLA has a core of professional militaries, diplomats and media and PR staff, including online presences of the MNLA and its own newspaper, Toumast Press. As with the Afghan powerbrokers, there is an ethnic Tuareg core, but as there are rival factions, the MNLA cannot exclude other ethnic groups and insists so in its own publications.

“The MNLA (National Movement for the Liberation of Azawad) would like to make it clear that within the MNLA military command there are: old rebels from the uprisings of the 1990s (MFUA - Movements of the united Fronts of Azawad), of 2006 (MTNM - The Tuareg Movement of Northern Mali, which was led by the late Ibrahim Ag Bahanga), fighters who have returned from Libya but who mostly participated in the liberation of that country, volunteers from the various ethnicities of northern Mali (Tuareg, Songhai, Peul and Moor) and both soldiers and officers who have deserted from the Malian army” (MNLA, 2012).

Tuareg powerbroker organisations were built on top of a clan and tribal structure to defend self-government and independent lifestyle. In exile, between rebellions, the clans formed modern organisations to better advance their cause after defeat in the 1963 rebellion.

c) Alliance behaviour

Pezard and Shurkin (2013, p. 4) find “[u]prisings are almost always the work of a few specific clans or tribes acting in pursuit of specific objectives, which tend to have a great deal to do with their position relative to other clans.” As the rebellions have raged through the north of Mali, state power has eroded. Mali is barely able to maintain a presence in the north and only with international support. “Mali is an anarchic state. Therefore, we have gathered a national liberation movement to put in an army capable of securing our land and an executive office capable of forming democratic institutions” (MNLA spokesperson, quoted by Mohanty, 2018). Frictions among clans and with other powerbrokers and the Malian state cause the dynamic anarchic environment in which survival and power matter more than ethnicity or religion.

One central figure to have analysed this is Ilyad Ag Ghali. His father was a leader of the 1963 rebellion and executed by other Tuareg for switching sides (Malijet, 2016). Ilyad Ag Ghali, a Kel Adagh of the Iriyaken trained in Libya, initiated the 1990 and 2006 rebellions. He initially led the Popular Movement for the Liberation of Azawad (MPLA). In 2011, Ag Ghaly lost the leadership struggle with other clan leaders. Ibrahim Ag Bahanga received support from Bamako during the 2006 rebellion. Bamako exploited divisions within the Tuareg. Bahanga joined the anti-Malian rebellion upon return from Libya in 2011 and relegated Ag Ghaly. Ag Ghaly took his tribe and formed Ansar Dine, recruiting from among his Kel Adagh and disgruntled Ifoghas. Ansar Dine allied with Al-Qaeda in the Islamic Maghreb (AQIM). AQIM is largely Arab and includes a small Algerian Salafist leadership group. The anti-AQIM Front de Libération Nationale d’Azawad (FLNA) is also mostly Arab, and once more a split house.

In early 2012, the MNLA, Ansar Dine and AQIM led a rebellion against Mali. The Islamist connection between Ansar Dine and AQIM and their rapid territorial gains sparked French intervention. When MNLA, Ansar Dine and AQIM jointly took the cities of Gao and Timbuktu, Ansar Dine and AQIM turned against the MNLA and pushed them out of the city. Realising the threat from Ansar Dine and AQIM, the MNLA started negotiating with Mali and France. The Songhai and Peul had their own self-defence militias (Ganda Koy and Ganda Iso). Whereas they attacked Malia army outposts in 2011 (McGregor, 2012), they allied with Mali as the combined Tuareg advances seemed the looming threat in 2012 (Chauzal and van Damme, 2015, p. 41). Sensing the shifting distribution of power in 2013, the Islamic Movement of Azawad split from Ansar Dine and leader, Alghabass Ag Intalla, announced he would renounce all violent Islamism (BBC, 2013). Intalla is the son of the amenokal who was passive in the 1960s rebellion, which was led by Ag Ghaly’s father. Intalla was rewarded for breaking with Ag Ghaly by becoming amenokal in 2014 (RFI, 2014).


Astructural realist approach appears useful. The powerbrokers discussed act as above the law and therefore induce anarchy and allow realist assumptions and thus theory to analyse the structure of the system. Alliance behaviour can therefore be described through tan analysis of relative power distribution, in this case, the Balance of Threat theory of Stephen Walt.

Afghan, Lebanese and Malian powerbrokers all confirm the hypotheses. The powerbrokers originate in a context of self-governing communities trying to maintain their internal autonomy vis-a-vis a more centralised state and world system. During conflict, the communities band together against a perceived external threat, building neo-feudal political-military bodies. Pooling military resources under skilled leadership and privileged access to outside sponsors sparks the birth of a post-conflict political-military elite. It is beneficial to maintain powerbrokers as they are better suited to distribute resources through patronage at scale. As the use of force is with the powerbrokers, the system remains anarchic and induces competition. Alliances are built with BoT logic. Green indicates full confirmation, whereas the yellow colour in Lebanon acknowledges that the LAF has maintained some independence and not all powerbrokers’ militias could be identified in this article, even though Hezbollah did confirm the hypothesis.

Table 4

Evaluation of case studies and hypotheses.

Self-governancePol-Mil NexusBoT factor
AfghanistanVillage, Clan, TribeArbaki, Mahaz, Tanzim, ANDSF influenced by powerbrokersAggregate power, aggressiveness
LebanonClan, Zu’amaClan Militia, Parties: Hezbollah, Amal, Phalange, FPM, Kataeb, LF, Future Movement; security influenced by powerbrokersPopulation, proximity
Northern MaliClan, TribeSelf-defence militia; MNLA, Ansar Dine, Ganda Koy, AQIMOffensive power, aggressiveness

Understanding powerbroker dynamics is especially important for political and peace-keeping missions. Elites in these systems are adapt at instrumentalising foreigners. Afghans have managed British, Russian, Soviet and US political rents for generations with no end in sight. Some have managed to convince US intelligence that their local enemies are Taliban. Lebanon manages to receive US, EU, Saudi, Iranian and other support at the same time.

External pressure can inadvertently put foreigners in opposition to the powerbroker system. Where two centralising visions are in conflict, e.g. westernised modernisers and Islamists, traditional powerbrokers can either be utilised if their interests are grasped or accidentally antagonised, increasing the severity and duration of the conflict. If not understood, powerbrokers might frustrate efforts by slowing down modernisers or even breaking with the consensus and oppose the state and international efforts openly.

Further research is needed to understand the extent of the phenomenon. Prima facie, it seems that a wide range of political systems could be understood by this approach. Central Asia, Libya, Iraq, Somalia and Sudan are the most obvious examples, but also the nations dominated by strong military states seemingly in perpetual conflict with their own societies, such as Iran, Egypt, Algeria and Pakistan, might be better understood by in-depth explanations of their political system with a structural understanding and a BoT logic. However, the distinction between pre-industrial kinship-based societies and post-industrial atomised societies is fluid and without a clearer and better developed definition and measurement between the two, a final tally is impossible and cases can become ambiguous. The theoretical implications are also not yet fully developed. Different strategies should apply for unipolar, bipolar and multipolar systems. Identifying these scenarios and comparing behaviour appears a fruitful endeavour to predict future conflicts.