Urban Resources and Their Linkage to Political Agendas for Armed Groups in Cities

Rapid urbanisation in countries undergoing or recovering from armed conflict has imposed severe strain on public services and governance. It has also been accompanied by increasing policy and academic concern about the vulnerability of cities to armed conflict. But even when major armed clashes between conflict parties concentrate in rural areas or towns, large cities play a critical but often less visible role: that of hubs for illicit economies supporting the political agendas of armed groups such as militias and insurgents. Urbanisation has reinforced the economic functions of cities for non-state armed groups willing to exploit the rising demand for services, amid weak state governance. This article argues that cities located near armed conflicts contain economic opportunities and incentives, linked to scarce public services amid rising populations, that are attractive to armed groups for their material and political value. These resources also provide incentives for the use of violence. The article develops the concept of urban resources, defined as sources of income for armed groups linked to the agglomeration of people and the scarcity of essential goods and services (housing, water, security provision etc.) resulting from inefficient urban governance. It uses the case studies of Mogadishu, in Somalia, and Karachi, in Pakistan. The article identifies two main categories of urban resources for armed groups: rents linked to the concentration of people within areas of weak state presence (extortion, charges for security services) and fixed assets linked to the growing demand for land and housing. It then examines the political benefits that have accompanied the exploitation of these resources.


Introduction
Rapid urban population growth in many developing regions (Sub-Saharan Africa, South Asia, parts of Latin America) has challenged governments' capacity to provide services and strained cities' infrastructure, housing and job markets (UN DESA 2018). This rapid demographic shift has also led to growing policy and expert attention to urban violence. Military planners have raised attention to the growing likelihood of urban warfare (Evans 2016). Latin America, which is highly urbanised in relation to other developing regions, consistently dominates the annual ranking of the most violent cities in the world (BBC Mundo 2018).
Views on the implications of urbanisation for low-income countries have often assumed that the expansion of slums and rapid population growth would bring about violence and anarchy: the dystopian 'feral cities' where 'the rule of law has long been replaced by near anarchy in which the only security available is that which is attained through brute power' (Norton 2003: 1). Robert Kaplan's 1994 famously pessimistic forecast of the ' coming anarchy' in developing regions expresses a similar point: 'Whereas rural poverty is age-old and almost a "normal" part of the social fabric, urban poverty is socially destabilizing' (Kaplan 1994). This sweeping view of ever-violent 'third world' cities bypasses the varied forms and pathways leading to urban violence (as this special edition illustrates)-after all, not all developing-world cities or slums are violent. Armed violence, crime and the presence of non-state armed groups are indeed serious challenges facing many cities, but these threats are sustained by underlying-and less headline-grabbing-flaws in the fabric of societies and political systems.
These underlying flaws are exemplified by the role of large cities, including capitals, in ongoing armed conflicts. The largest population centres in Somalia and Afghanistan (Mogadishu and Kabul, respectively) intimidate and sometimes control low-income areas, which tend to concentrate illicit economies and lack rule of law. Violence for control or supremacy over low-income settlements where urban resources cluster, therefore, has developed dual economic and political functions.
This article draws from a total of 30 interviews conducted during field research in October and December 2019 in Mogadishu and Karachi and is part of a project entitled 'Urban Drivers of Political Violence', which has looked at non-state armed groups and violence in large cities located near armed conflict. Several interlocutors are not identified by name for security reasons. Interviews focused on current and former state employees (especially those in the security sector), members of international organisations, local experts and members of civil society organisations (such as non-governmental organisations). Secondary sources have also been used in this article. The next section lays out in more detail the concept of urban resources in conflict-affected settings, with reference to the broader literature on the political economy of conflict and non-state armed groups. The article then proceeds to the two case studies, followed by a discussion in which the types of urban resources are identified, alongside their economic and political benefits for armed groups.

Urban Resources and Hybrid Economic-Political Violence
Cities concentrate 55% of the world's population and 80% of global gross domestic product (GDP), according to The World Bank (2020). Countries affected by armed conflict are no exception to this trend, and in fact the concentration of military forces in large-usually capital-cities encourage further concentration of domestic economic activity as well as the investments by international aid organisations (which base themselves and employ mostly in such cities). The average urban population growth for fragile and conflictaffected countries in 2019 was 3.2%, almost double the global average, which was 1.8%. 1 Yet, the bulk of the literature on the political economy of armed conflict is focused on natural resources. The political economy of conflict became a major point of academic contention following Paul Collier's argument on rebels' 'greed' for natural resources (meaning primary commodity exports, such as diamonds, timber and coca for drug production) as 'the largest single influence on the risk of conflict' (Collier and Hoeffler 2000: 26).
The picture of financing for conflict actors is a lot more complex than that, with natural resources serving important roles in some cases while, in others, combatants focus on local or transnational organised crime and even on revenue from investment in legal business sectors (Wennmann 2007: 432). Studies have shed light on the wide range of roles and pathways through which economic factors (meaning sources of income for armed actors) affect armed conflict. For instance, economic agendas may gradually gain prominence as a motivation for violent armed groups in conflicts initially ' driven by military and political considerations' (Berdal and Keen 1997: 798).
Importantly for our analysis of the economic functions of cities for conflict are the findings related to the connection between economic 'greed' and political 'grievances'. Corruption and inequitable distribution of resource wealth can themselves be grievances fuelling armed conflict (Malone and Nitzscheke 2005: 5); whereas, control over illicit economies has provided non-state armed groups with ' considerable social and political capital' in areas of poverty and weak state presence that rely, for instance, on coca and opium as sources of income and jobs (von Einsiedel 2017: 5). Awareness of the different effects of armed groups' economic activities on political dynamics is crucial for the analysis of urban conflict, because any rebel economic endeavour in a large urban area will inevitably interact with larger population densities and more complex social and political geographies than those in rural areas. Economic exploitation of urban resources requires regular interaction and regulation of human communities-often in densely inhabited low-income settlements-as economic value is derived from human and capital resources, such as taxation on trade, provision of security and land dealings.
If recent authors on the political economy of conflict have made cities marginal to their analyses, one of their greatest predecessors saw urban resources as a major factor in European warfare and state formation. Charles Tilly argued that the different forms of social organisation in European territories 'significantly affected the strategies rulers employed to extract resources', with some contexts and groups offering more resistance or requiring more compensation than others (Tilly 1992: 27). Tilly highlights two prominent models of resource extraction for war: in rural areas, coercion was more commonly used by rulers; in urban areas, markets, commercial exchange and other capital-generating factors played a larger role in the relationship between rulers and social groups (Tilly 1992: 18). In the urban context, Tilly describes a model of 'fragmented sovereignty' marked by ' compacts with capitalists-whose interests [rulers] served with care-to rent or purchase military force, and thereby warred without building vast permanent state structures' (Tilly 1992: 30). Tilly's theory of state formation through warfare-and the resource extraction models required for ithighlights, therefore, the propensity of urban resource extraction to acquire the format of alliances between the capital-oriented groups rather than the direct application of military coercion prevailing in rural areas.
The process of ' extracting resources' for war, far from being universally a literal process of directly accessing primary commodities (though sometimes it is), involves ' alliances with specific social classes'. Even though Tilly makes these points as part of his analyses of war and state formation, he highlights that these relationships for resource extraction were performed by a variety of 'power holders'-indeed he alludes to them as types of organised crime-to ' enjoy the advantages of power within a secure or expanding territory' not usually intentionally building national states through such extraction (Tilly 1985: 172).
Tilly's insights into resource extraction in the process of war and state formation provide useful points for our analysis of armed groups' resource extraction in cities. Many armed groups exploit urban economies while openly claiming to be state makers-insurgents and rebels such as the ones in Karachi and Mogadishu. But even if the process of expansion to large urban areas was not part of a grand strategy to conquer such cities (which seems like an unrealistic prospect in Mogadishu and Karachi), it is important to bear in mind that resource extraction has been conducted by a wide range of armed actors, some possessing legitimate claims over the use of force and others not so much.
Urban resource extraction by non-state armed groups in contemporary conflicts takes place in a context of weak state capacity amid rapid expansion of urban areas and populations. This process facilitates the encroachment of armed competitors to the state into major urban centres to extract resources even in proximity to states' military assets: al-Shabaab has been able to establish a stable extortion mechanism in Bakara Market, in the same city where the bulk of national and international forces are concentrated. This resource extraction is more discreet than in rural areas, where both al-Shabaab and the Pakistani Taliban established much more direct territorial control through visible military force. This process resembles what Tilly (1992: 15) called 'fragmented sovereignty' in cities, with the wielders of military force leaning towards alliances with holders of capital in cities as opposed to ' coercive-intensive' rural areas.
Alliances and co-optation are also prominent mechanisms for the ' economic functions' of violence in internal wars described by Keen. While contrasting the views about armed conflicts as primarily political or military struggles, Keen (1998: 12) suggests that 'rather than simply asking which groups "support" a rebellion or counter-insurgency, it is important to ask which groups take advantage of these situations for their own purposes'. The ' advantages' of conflict economies are enjoyed not only by armed groups extracting resources but also by a range of 'legal' actors who may exploit conflict for direct financial gains as well as 'the pursuit of special privileges', including government officials (who may divert military or aid money for personal gains), landed elites (who may see conflict as a way of preventing land reform) and traders profiting from weaker border restrictions (Berdal and Keen 1997).
But while these economic relationships are conducted alongside the use of coercion and direct territorial control through military means in many rural areas, armed groups face major obstacles to publicly claim territorial control in cities-mainly due to the greater proximity to national and international military forces. That, alongside the concentration of diverse and often rival social, ethnic and sectarian groups in cities, propels non-state armed groups towards alliances and co-optation alongside violence (or the threat of it). The city's economic functions for armed groups, therefore, are intertwined with political relationships with urban groups. The case studies below will exemplify that by describing al-Shabaab's somewhat ' discreet' coercion techniques towards traders in Bakara Market and the Pakistani Taliban's exploitation of disenfranchised Pashtun peripheries of Karachi.

Political Fragmentation and Economic Opportunities
The fragmentation of political authority in cities has been a major trend in various contemporary instances of organised armed violence. Rapid and unmanaged urbanisation means that state institutions struggle to provide services, legal economic opportunities and security in vast peripheries and slums. This is especially the case in countries undergoing armed conflict or recovering from it, where urban governments and infrastructure have been further strained by inflows of displaced populations seeking safety and jobs. Although this urbanisation process has not led to outright anarchy and cities are often considered safer by many that witnessed fighting in rural areas, it has created some opportunities for non-state armed groups.
Non-state armed presence in cities, alongside the heightened security measures in cities when compared to rural areas, make cities particularly prone to 'multi-layered governance': not a binary split between rebeland state-controlled areas but 'various armed groups opposing states (…) as well as cooperating with or opposing each other' (Kasfir, Frerks & Terpstra 2017: 259). In such situations, areas within cities can witness 'parallel power systems' imposed by a variety of armed groups-not only well-organised insurgents but also ' community-based armed groups' guided by parochial purposes, such as support to local strongmen or hiring as armed wings of political parties (Schuberth 2015: 304). The connection to influential political parties or figures does not necessarily comprise a grand ideological or revolutionary purpose but is mixed with profit-seeking and control of economic assets, such as land, water mafias and extortion (Schuberth 2015: 298). This blurring of political and economic motivations is exemplified below in the case of political parties' armed wings in Karachi, which have fought each other for territory while also controlling extortion and other illicit revenue streams tied to the control over space. In the context of Somalia, Ken Menkhaus observes that 'warfare has been replaced by tense stand-off and divided communities -neither war nor peace ' (2004: 41), which helps explain the availability of groups akin to militias for hire to influential individuals grabbing land in Mogadishu.
This splintering of political authority among state and non-state actors is affected by the political geography of cities in conflict-affected countries, particularly the way that political power is exercised at different levels 'in the face of the failure of local governments to meet growing urban needs' (Lindell 2008(Lindell : 1879(Lindell , 1895. A distinguishing factor of urban armed conflict in Sub-Saharan Africa, for instance, is the way in which violence and political contests from broader (national or regional) conflicts are '"spatially" translated, resulting in (ethnically) fragmented urban neighbourhoods or violent struggles over urban public space', in which 'the urban material space is often actively used in strategies of different actors involved' (Büscher 2018: 199). Violence from broader national or regional armed conflicts assumes specifically urban characteristics once conflict actors reach cities (Sampaio 2019).
The clustering of people and businesses amid scarcity of formal service provision (through the state or private sector) feeds economic opportunities tied to extortion and rents. The demand for these services, and above all for security, is met by informal providers, something that in contexts as varied as Medellín in Colombia and Karachi in Pakistan has served as a useful entry point for insurgent groups linked to nearby armed conflicts to increase their ties to low-income areas (Sampaio 2019).
A key economic function of informal areas (slums and other unplanned peripheries) consists of 'the physical concentration of dangerous illegal activities' due to the predominance of informal employment (without formal labour contracts and social security) alongside myriad opportunities in an 'illicit world of violence and impunity' (Davis 2009: 231-232). Clustering of population, businesses, informal employment and illicit economies, therefore, provides incentives for armed groups with political and ideological agendas to become involved in illicit economies.
Another important function of cities close to armed conflict for non-state armed groups relates to the concentration of financial and political capital. With the inflow of displaced populations, alongside economic migration and organic growth, large cities such as Mogadishu, Karachi and Kabul have grown rapidly and without planning, leading to rising prices for land and housing. This market has become a growing source of interest and revenue for militias and insurgents, which have acquired roles in grabbing, regulating and reselling urban land. Furthermore, urban land conflict tends to involve politically contentious issues, such as the entitlement of certain ethnic groups to certain lands, rights linked to 'indigeneity' or history and other inter-group rivalries that immediately render an ' emotional dimension' to fights over a seemingly economic asset (Lombard andRakodi 2016: 2687).
Similar entanglement between land disputes and simmering inter-group political rivalries has fuelled violence in Karachi between 2008-2015, with political parties and their militias, often organised along ethnic lines, acting alongside developers and corrupt police officers to evict people and murder others who resisted. 2 The usefulness of urban land for political parties included a combination of monetary value and political capital among their political constituencies and ethnic groups. The fight for territory between rival political parties through their militias involved the protection to each party's ethnic constituency from rival groups. But this also meant that 'the resources belonged to the armed group in control'. 3 Invariably, land was among the most valuable resources, because it would be subdivided and sold, but also controlled according to the political and ethnic interests of the group in place. 4 Insurgent groups, despite their connection to religious causes and armed conflicts against national governments, have become involved in urban land conflicts in both of the case studies below. The economic functions of the two cities reoriented at least part of the Taliban and al-Shabaab's energies from grand political rebellions to 'parochial' local interests tied to the political economy of the city (Schuberth 2015: 298). In Karachi, the Taliban took over the relationships with land mafias after wresting control over some Pashtun peripheral areas from the Awami National Party (ANP). In Mogadishu, al-Shabaab has developed an important role in resolving the growing number of disputes over land ownership, given the group's reputation for being much faster and more ruthless in enforcing its decisions than the state. 5 These economic assets of cities for armed groups result neither in a purely criminal enterprise nor in a carefully concocted plan to fund insurgencies for purely political purposes. In the two cities studied below, armed groups' control over urban resources almost always involved ties to non-ideological criminal actors, such as gangs, as well as to political ones, such as corrupt governments, parties or security forces.
The usefulness of urban operations by armed groups has constantly swung between political and economic functions. Economic and 'grievance' motivations become tied together: even if economic factors were not prominent motivations at the onset of armed conflicts, they become increasingly critical drivers 'in the persistence of open fighting, localised violence and coercion ' (Eaton et al. 2019: iv). Urban resources play a critical role in attracting and sustaining armed actors linked originally to 'grievances' over political or ethnic causes, facilitating and providing incentives for the gradual incorporation of localised control over urban rents to violent groups' core activities.
But urban resources, like other types of economic resources for insurgent organisations (including natural resources as well as external state or diaspora support), do not dictate an outright ' degeneration of armed groups' towards criminality nor a tightly obedient political-ideological movement (Staniland 2012: 142). Rather, the theory developed in this article considers urban resources as sources of income that necessarily entangle armed groups in negotiations and bargaining with diverse social groups inhabiting an urban area, often in close proximity to each other-be them oriented along ethnic, religious, economic (as in classes such as traders, land owners etc), party, criminal or other lines. Different insurgents will see different effects of urban resources for their organisations-as in Paul Staniland's analysis of how the strength of social ties within an insurgent group is 'the crucial conditioning variable' on how resources relate to insurgent organisation (Staniland 2012: 174). In other words, urban resources do not necessarily make insurgents become common criminals; if anything, they require at least some political messaging in order to advance their cause as part of relations with urban communities.
The preceding analysis on urban resources for political violence, as well as the following case studies, provide an important addition to discussions of the political economy of conflict (and sub-economies of specific geographic areas), which have nuanced the old discussion over whether conflict actors are motivated by 'greed' or 'grievances' (Eaton et al. 2019: v).

Mogadishu Case Study
The Somali capital is significantly less affected by armed violence today than during the 1990s and 2000s, when the collapse of central government placed the city in the centre of fights between clan militias, independent Islamic Courts (later grouped In the Islamic Courts Union, or ICU) oriented around sharia law and a period of control by Islamist insurgent group al-Shabaab. Beneath the surface, and despite the fact that it is the main base for national and international military forces, Mogadishu has been a centre of competing claims for sources of revenue, including by al-Shabaab but also by less-organised interests linked to wealthy landlords with access to private security forces. The political control that non-state armed groups had in previous decades over territories in Mogadishu has given way to an economic control: for instance, in Bakara Market, the city's main commercial area and probably one of the busiest areas in Somalia, al-Shabaab is able to extract regular extortion revenue from businesses. 6 Significant tracts of land in a city with a rapidlyexpanding population have been controlled by wealthy individuals with sporadic use of armed support. 7

Al-Shabaab's resource extraction in Mogadishu
The militant Islamist group Harakat al-Shabaab al-Mujahideen lost control over major urban centres in 2011 and 2012. The group made what it called a 'tactical withdrawal' from Mogadishu in 2011 and was expelled by African Union forces from the country's second-largest port, Kismayo, in 2012(BBC News 2018. Despite now being in control of mostly rural (albeit expansive) areas in southern Somalia, the group has a wellorganised support network in Mogadishu that provides intelligence, infiltrates security forces, conducts devastating terrorist attacks and extracts extortion from businesses. Its ability to operate in the capital, especially outside the high-security area around Mogadishu International Airport (MIA), derives also from the weak institutions, especially the security sector (particularly the police and the army) and the courts. On top of these institutional flaws, the longstanding weakness of the national state since the collapse of President Siad Barre's regime in 1991 has led clans and business leaders to pursue their own interests and sometimes clash with each other through militias and other privatised security forces (Menkhaus 2004: 41-42).
This combination of state weakness and parallel sources of political authority is reflected in the country's political centre, Mogadishu. Despite al-Shabaab's 'tactical retreat' from the capital in 2011, its ability to continually strike and impose taxation (or extortion) from businesses there fulfil important symbolic and practical purposes. Whereas interlocutors have told this author that al-Shabaab has indeed ' criminal' activities in the city (mostly extortion), several have also outlined a political role for this fundraising activity: not just the strengthening of Somalia's most powerful insurgent group but also the way this activity damages government authority. For instance, a leader of a local civil society group told us that ' community leaders and al-Shabaab have a dialogue, you can't refuse what al-Shabaab demands'. 8 Government authorities interviewed during field research in the city have recognised that al-Shabaab is able to extort money from people and businesses but said this was evidence of the group 'transforming itself into a mafia organisation, maybe not as sophisticated as the Italian mafia, but the way it behaves is not ideological anymore'. 9 However, sources in international organisations and the expert community have a different take: fundraising through criminal activities in Mogadishu and elsewhere is strongly connected to the political goals of the broader organisation. On the one hand, a member of an international organisation keeping a close eye at terrorism finances in Somalia said 'the phenomenon of extortion in Mogadishu is pure criminal behaviour', but at the same time 'they fund their ideological drive through criminal activity'. 10 The two factors-economic and political motivations-seem to be connected (as stated in our previous sections): a UN Security Council Panel of Experts report mentions al-Shabaab's taxation of imports coming into Mogadishu port as an example of the group's expanding revenue-generation ability and a source of 'resilience in the face of increased security operations' by national and international forces (Bahadur et al. 2019: 3). Furthermore, the control over traders and activity in Bakara Market has produced some military benefits: al-Shabaab has set price limits on certain small arms being sold there during a period in which their prices were soaring (Graduate Institute of International and Development Studies 2013: 269). 11 Another aspect that makes urban resource extraction not just ' criminal' but also politically consequential is the systematic manner through which extortion is conducted; in other words, it is not just an ad-hoc inconvenience but a well-organised and regular extraction of money that al-Shabaab considers zakah, or almsgiving, that the group imposes on business people to pay at 2.5% of the monetary value of a business before profits (Hiraal Institute 2018: 2). The way this tax is collected displays the group's ability to maintain a regular operation in the city through a combination of physical threat and remote influence: business leaders are often contacted via calls to their mobile phones and summoned to tax collection offices located in nearby villages (Faruk 2019). Al-Shabaab supports this system through a capable information collection capability: business leaders have revealed in a recent Washington Post report how the group had detailed information about trade containers even before they reached the city (Faruk 2019). Furthermore, al-Shabaab uses informants in the city to estimate the volume of goods maintained and received by businesses, whose owners are then contacted to pay the tax and, in case of refusal, are reported to its intelligence wing, Amniyaat, for further pressure or potential punishment (Hiraal Institute 2018: 4).
The impact of this financial activity does not stay restricted to the economic realm but reverberates through other areas of life, such as informal rules of behaviour. For instance, a local resident said she pays particular attention to her dress when she visits Bakara Market because of the well-known al-Shabaab influence there. Furthermore, whereas the money collection may be conducted through remote communication technology, the enforcement aspect is not. An Interior Ministry official said al-Shabaab's enforcement system creates real-world fear: [Al-Shabaab members] call and demand money and threaten the business owners, if they don't pay, to come and kill them, [saying] that they are everywhere. When they do the punishment, they gather all the people to see it, they create fear. People believe al-Shabaab is more powerful than it really is because of the harsh punishments and bomb attacks that create fear. Fear is greater than their capacity. That fear is important for them to implement their agenda. 12 This fear also spreads outside the business community. The market is considered off-limits to government staff and foreigners. 13 A local resident said a friend who works for a United Nations agency was shot dead while venturing in the market, with al-Shabaab being a prime suspect of the murder. 14 Despite the predatory character of extortion and intimidation, this activity fits within a broader narrative by al-Shabaab as a source of political authority able to rival the government, and sometimes surpass it, in provision of public services. For instance, roads controlled by al-Shabaab are preferred to those controlled by the government: When people leave the city to go somewhere else, they prefer to use al-Shabaab routes because the government routes will have charges in each checkpoint, whereas in al-Shabaab roads you only pay once and that payment will last through the rest of your journey. 15

Benefits of insecure land tenure for non-state armed groups
Al-Shabaab also fills an important institutional gap for a city where disputes over land tenure have been growing in recent years. The group provides quick and efficient dispute resolution through ' courts' located near the city. Residents have the additional option of resorting to government courts and clan elders' mediation, which take longer and lack effective enforcement mechanisms for their decisions. Not only are many Somalis not accustomed to going to government agencies for dispute resolution, but also the courts can take up to a year to arrive at a verdict; whereas, a government interviewee said al-Shabaab courts can take two to three days. 16 As in the case of extortion, al-Shabaab's use of violence serves as a backup: insurgents will sometimes call the losing side of an argument and order land to be vacated 'within hours', with punishment being harsher and more certain than in the case of government or clan decisions. 17 This divergence between the different dispute resolution mechanisms has been acutely felt since 2011, as Somalis began to return from abroad and find their land and houses occupied by others. 18 Conflicts related to land, ranging from ownership to forced evictions, have been major source of violence in Somalia. This goes beyond al-Shabaab and involves deep-rooted rivalries between clans fighting for control of lucrative assets linked to land since the collapse of the central state in 1991-including rural resources such as pastures for cattle, riverine agricultural areas and, increasingly, valuable urban places, such as airports, seaports and real estate markets (Menkhaus 2018). Even though clan militias have been nominally incorporated into the official security services and clashes between rival factions have almost completely disappeared from Mogadishu in recent years, wealthy or well-connected landlords are still able to access armed men to acquire or maintain land properties.
These armed men represent what one interviewee called 'the dormant fighter' militias: even though they are no longer part of an organised clan militia, they are activated to forcefully remove people from land or to fight off rival claimants-sometimes due to loyalty to a certain clan faction but also sometimes due to people 12  'in positions of power' who can order members of the police force to do their bidding despite the conflict with public interest. 19 This was corroborated by Dr Hodan Ali, manager of the Durable Solutions Unit, a local government agency supporting displaced populations in Mogadishu and its surrounding areas: people wishing to enforce their claim to a certain land 'force people by using an armed apparatus', which includes members of the police force. 20 She cites a case of mass displacement in December 2017 when thousands of people were forced out of displacement camps in the outskirts of Mogadishu through the use of 'militias' and bulldozers that even destroyed structures built by humanitarian organisations (UN Somalia 2019: 12). According to Dr Ali, 'the person who wanted the land back just called some men through the police and military and asked them to get people out'. The state security sector in Somalia is so weak, and the idea of pursuing private interests through force so entrenched, that some of the police didn't even bother to take their uniforms off while working for the aggrieved landlord. 21 Weak urban governance, with the state competing or co-existing with these private interests, facilitates the violent pursuit of financial interests linked to land. The 'lack of a clear system to recognize and uphold land ownership and user rights', added to the looting or destruction of public cadastres in recent decades, has resulted in frequent competition (UN Somalia 2019: 12). In a city where militia-like structures are still available for private use, weak land tenure structures have been turned into financial opportunities by claimants willing to use force.
Whereas land is an intrinsically valuable asset in any urbanising country, in Somalia this trend is linked to armed conflict dynamics in two ways. First, the aforementioned weakness in the security sector and the availability of former militia members still available and willing to pursue private interests-the ' dormant fighter' militias, a legacy of the wars since the 1991 fall of the Siad Barre regime-facilitates the violent exploitation of land interests. Second, Mogadishu has faced extremely large population inflows even for the standards of Sub-Saharan Africa (one of the most rapidly urbanising regions on Earth). Somali cities currently host just over two million people displaced from their original homes, not only due to armed conflict but also due to drought and other climate-related issues (UN OCHA 2019: 6).
This combination of factors-demographic pressure, weak governance and availability of non-state armed groups-converge in a single urban area already overwhelmed by fragile state institutions and displacement from nearby conflict areas. The availability of urban resources amid weak state capacity have facilitated their exploitation by armed groups with both economic and political benefits.

Karachi Case Study
Karachi, the world's 12th largest city, has been affected over decades by its position close to prolonged armed conflicts, mainly in Afghanistan and Pakistan's north-western tribal areas (UN DESA 2019: 29). Populations fleeing conflict and many other migrants have flocked to the city, Pakistan's main port and industrial centre, in search of economic opportunities. As a result, Karachi grew by 130% between 1988 and 2018, reaching 15 million inhabitants. 22 Like many other large cities in South Asia, a huge section of this urban growth has been concentrated in informal areas, where 55% of the city's population live (UNDP 2018: 7). The increasing population has been accompanied by rising tensions, as rival political parties and ethnic groups competeoften violently-for influence, votes and profitable stakes in the city's informal economies, including land grabbing, illicit water and electricity provision.
The economic opportunities were manifold: a huge demand for services not adequately provided by the state, a thriving land mafia business and often corrupt state institutions. The competition for these urban resources attracted actors endemic to the city as well as external groups from nearby armed conflict theatres. During the height of political violence in the city, between approximately 2008 and 2015, armed groups vying for economic gains and territories included militias affiliated with political parties, gangs and insurgent groups, especially Tehrik-e-Taliban Pakistan (TTP, the main Pakistani faction of the radical Islamist group Taliban). This case study focuses on this period of heightened violence, after which a military operation led by Army Rangers started producing results in quelling many of the ethnic and sectarian clashes.

Political militias vie for resources
Rapid urban population growth has had a role in intensifying competition between rival political and social groups due to the rising pressure on public services and disruption of the city's tenuous equilibrium of ethnic and sectarian groups. Pashtuns from Afghanistan and north-western Pakistan (Khyber Pakhtunkhwa province and the Federally Administered Tribal Areas, FATA, which were merged in 2018) made up the majority of the displaced population fleeing conflict towards Karachi since 1979 (when Soviet Union troops entered Afghanistan to shore up an allied regime). Several ethnic riots broke out in the late 1980s (especially 1985-1988) as the influential Muttahida Qaumi Movement (MQM), a political party representing the Mohajirs-or migrants from the partition of India and their descendants-feared the rising political impact of the city's demographic shift and distributed leaflets against Pashtuns in their strongholds (Ur Rehman 2017: 69). This demographic shift also altered the balance in the city's informal settlements, many of which were managed by Mohajir entrepreneurs but became increasingly disputed as Pashtun migrants also flocked to informal areas that were far more affordable than Karachi's 'formal' and more affluent areas (Yusuf 2012: 6).
Control of profitable income from land in these informal areas has been at the centre of the many bouts of armed violence in the city perpetrated by rival political militias. The unregulated and unplanned nature of these vast informal areas (known locally as katchi abadis) meant that violent armed groups, usually land mafias and militias, stepped in to grab land and offer associated public services. These illicit armed actors operating around land became embroiled in the city's violent political rivalries as parties would seek control over informal areas, often through alliances with land mafias, in order to provide housing and some services to their constituents and therefore enhance their political influence (Yusuf 2012: 12). Insecurity over the status of land and recurring violence along ethnic and political affiliations contributed to areas acquiring an increasingly homogeneous ethnic composition and the splintering of the city into separate ' citadels' comprising what Laurent Gayer (2014: 41) calls the dissolution of the city's cosmopolitan character despite its hugely diverse social mix. This fragmentation of governance over chunks of Karachi across armed factions ended up connecting the fight over political power to that over control of illicit economic revenues, especially urban land.
Illicit economic interests gained central stage for armed groups when violence began to escalate again from 2008 onwards. This new bout of violence was linked to disagreements and jockeying for territories between three main political parties (the MQM; the Awami National Party, ANP, which represents Pashtuns; and the Sindh-based Pakistan People's Party, PPP) and their armed wings (Yusuf 2012: 7). As the rivalry around political influence flared in several urban areas, especially katchi abadis, control over economic resources gained even greater strategic importance as it became a further way of marking control over urban spaces against rival militias as well as an obvious target for rival armed groups. A local architect described the high economic stakes involved in the fight for territorial control: In each territory, the resources belonged to the armed group in control. Land available would be controlled by the ruling party, which wold subdivide it and sell it. There were also informal businesses, lots of shops etc, and they would always pay a rent-an informal rent-for protection to those political groups. 23 Whereas political militias had a strong interest and role in the control over informal settlements and their accompanying informal economies, these opportunities were also exploited by criminal gangs. The weak governance and legitimacy of the state in many areas of Karachi only reinforced local dwellers' reliance on non-state groups enforcing property contracts, regulating economic transactions and handling dispute resolution in the absence of the state (Gazdar & Mallah 2013: 3112). Organised crime, however, did not represent an entirely separate phenomenon, because a frequent method of such groups was the establishment of collusion with political parties. As Nazia Hussain and Luise Shelley (2016: 12) have argued, ' crime groups play an integral role in the political economy of the city' by serving as ' enforcers' to carry out violence against political opponents.
Political violence and resource extraction became intertwined in a way that was difficult to separate the two. An example is the creation of the People's Aman Committee in 2008 by a prominent gang leader in the large low-income area of Lyari, with support of the PPP amid a growing political rivalry with the MQM.
Around the same time, various groups introduced the collection of Bhatta-or extortion money-imitating a practice popularised by the MQM in other areas of the city (Kirmani 2017: 110). Control over Lyari became important not only because the area's 1.6 million dwellers constitute a sizeable political constituency but also because it is located near some of the city's most economically valuable areas, including the port (Kirmani 2015: 1, 6).
Armed groups' interests and disputes integrated economic and political assets, for instance, urban territory contained political constituencies and therefore votes in exchange for the party's role as local protector against rival ethnic or political groups. But this political asset-votes-was inseparable from the ability to regulate and control economic assets linked to public services not provided by the Pakistani state. This is visible in the distribution of economic benefits in exchange for political support in MQM-controlled areas, what Laurent Gayer calls ' armed clientelism'. In this MQM model, the party's armed wings provided not only physical security to dwellers in its territories but also 'muscle' for retaliations or intimidations against business rivals, in other words ' economic protection' (Gayer 2012: 78). Similarly, day-to-day activities of parties' armed wings-the political militias-involved routine involvement in criminal activities linked to resource extraction, including extortion (bhatta) and land grabbing (Gayer 2017: 92). Land grabbing (qabza) in particular 'served the electoral and larger strategic interests' of political parties (by enlarging vote banks) as well as presenting financial opportunities linked to housing developments (Gayer 2017: 97-98). In the larger scheme of local governance, public service agencies themselves were turned into 'resources' available for political parties as spoils for achieving electoral victories: both the PPP and the MQM have politicised institutions they controlled within the Sindh regional government and Karachi's local administration by providing services on partisan grounds to clients (ICG 2017).

Taliban's territorial and economic control
Competition over such rents became even fiercer around 2009, when Tehreek-e-Taliban members started moving to Karachi amid a crackdown by the Pakistani armed forces in their bases in Khyber Pakhtunkhwa (specifically the Swat region) and FATA's South Waziristan agency (ICG 2017). Given the Pashtun ethnicity of most of the group's members, Karachi's peripheral areas that had mushroomed in size due to years of displacement from conflict-affected Pashtun areas were taken over by the Taliban. The militants attacked their main competitor for political control over such areas, not the state itself but the ANP, the main Pashtun political party. 24 This territorial control was, from a very early stage, closely associated with the illicit economies in such areas. Local criminals previously associated with the ANP started providing information on locals to the Taliban for extortion purposes. 25 Another shift in criminal allegiances took place with land grabbing. 'Muscle men'-criminal groups-profiting from encroaching into land and reselling it needed the protection of a 'larger umbrella' group that provided connections to land agencies, which in Pashtun areas tended to be handled by the ANP. 26 But land grabbers 'found that the Taliban could be this larger umbrella after they arrived and attacked the ANP'. 27 In any case, the land mafias had little choice. Given the longstanding failure of government agencies in managing the land market, 'investment in land grabbing became the domain of those in charge of a certain area by virtue of their muscle and armed force'. 28 Several interviewees have said that there was no grand political ambition connected to the TTP's plan in relocating to Karachi, in terms of conquering territory or replicating the same type of power they had in more sparsely inhabited provinces. Rather, the TTP's objectives seemed to evolve gradually and to replicate models used by other Karachi-based non-state armed groups.
There was no military ambition behind [the TTP's expansion in Karachi]. They came here after safe haven. As the safe havens grew, they started building ghettos and areas of militarisation. They never wanted to take over Karachi but they certainly took over [certain] areas. Karachi is a large city with lots of disgruntled people, many of them Pashtuns, who are the most marginalised group here. They were able to take advantage of the obscurity of many areas of Karachi very easily. 29 This was the first megacity ever to host a significant presence of any Taliban faction, and as the group adapted to the local reality, the exploitation of economic opportunities tied to the city's population density grew. A former law enforcement official said the Taliban took part in land grabbing activities and used their Karachi areas as safe havens for smuggling Iranian diesel brought from Balochistan (a province bordering both Afghanistan and Sindh province). 30 Another important activity they replicated from other armed groups in the city was corruption of state agencies. The Taliban would make police agencies in some Pashtun peripheries their 'partners' if officers chose to 'turn a blind eye' and occasionally take an active role in smuggling activities. 31 The case of Karachi, with its combination of political militias, gangs and an insurgent group, indicate that 'ideology and crime don't operate in silos', as one local expert said. 32 Factions within secular organisations, such as the MQM, would at times cooperate with the Taliban, without this constituting a broader institutional linkage. 33 The involvement in the city's many illicit economies was facilitated by the precarious governance structures present in informal areas and the absence of any significant state presence. This criminal activity took place at the same time as the Taliban was expelling political rivals such as the ANP, instituting roadblocks to control access to areas and offering dispute resolution to locals, all essentially political activities that challenged or replaced state claims of governance. Resource extraction and political objectives, therefore, were intertwined and run in parallel for the Taliban, as they did for the political militias.

Discussion
The absence of rule of law and the presence of illicit economies have created particularly prominent opportunities for the financing of armed groups in Mogadishu and Karachi. Cities concentrate wealth and assets that are valuable but also vulnerable to violent takeover, particularly in countries with a recent history of armed conflict (Adawe & Menkhaus 2018: 31-32, 37). These assets, which I have called urban resources, include land, water distribution, security provision and taxation (often extortion) of businesses.
These resources are tied to spatial and social features unique to large cities, setting incentives and providing opportunities that guide the behaviour of armed groups (Weidmann 2009: 527). These features include, for instance, rapid population growth within a limited space, growing competition for land and the political importance of large cities in terms of votes and symbolic value. Furthermore, whereas weak service provision and other governance failures are not unique to urban areas, they produce in conflict-affected cities an ' exacerbation of spatial inequalities' (Raleigh 2015: 91). Urban resources, therefore, are tied to the scarcity of services and goods critical for low-income communities, particularly in conflict settings affected by rapid urbanisation and governance weakness.
There are two main types of resources arising from the spatial and social environment of conflict-affected cities, as observed in the two case studies: rents linked to the clustering of people within areas of weak state presence and fixed assets linked to the (growing) demand for land and housing in cities. Urban resources also have important political implications for non-state armed groups' behaviours and strategies: control over vital economic activities in spaces with such large populations and political importance have strengthened militias' links to certain ethnic groups and advanced insurgents' claims of political authority while weakening the state's legitimacy within the same urban areas. We explore the two urban resource types and their political implications below.

Types of urban resources for non-state armed actors
First, rents linked to population density include extortion and provision of rough services (often for exploitative prices). The extortion of local dwellers or small businesses can be considered-and indeed are called by some armed groups-taxes. These extortion rackets are often tied to the provision of security in marginalised areas where law enforcement is weak, absent or feared due to heavy-handed tactics, amounting to security taxes. In Mogadishu, for instance, several interviewees pointed to the government's failure in providing law enforcement, despite recognising that clan militias no longer openly patrol the streets. Mogadishu authorities ' don't have any strategy except the checkpoints' that dot the city and slow traffic, with the additional complication of al-Shabaab infiltration in the police. 34  state security providers, such as militia-like structures loyal to clan or corrupt government officials, some of which have operated protection and extortion rackets (Menkhaus 2016: 7, 24). Al-Shabaab's extortion practices against businesses also comprise a rough type of security tax, as the group is infamous for its ruthless enforcement of decisions made by what are called locally ' al-Shabaab courts', which people and businesses can and do use as a source of dispute resolution. 35 Another failure in urban governance that leads to rents for armed groups relates to the provision of a broader set of services, such as water, sanitation and electricity. The supply of water has been a particularly acute problem in Karachi, with criminal groups in association with political parties and their militias siphoning it off from the official grid and selling it through water tanks in marginalised areas where militias were in control (Hussain & Shelley 2016: 8). Other cities located in fragile settings have experienced similar problems. Water is among the key services provided in Kenyan capital Nairobi's Mathare slum by gangs, who benefit from weekly payments by each household-sometimes in alliances with water mafias with connections in government. 36 Second, conflicts over fixed assets, especially land, have been critically important in both cities studied. As in the case of extortion and security taxes, control over land provided a dual benefit for armed groups in urban areas: political authority over an increasingly scarce economic resource and income from land grabbing and reselling. Interviewees in Mogadishu have mentioned land eviction as one of the most prominent sources of insecurity in the city (alongside al-Shabaab terrorist attacks), carried out by 'militias' and ' dormant fighters'. 37 This type of non-state control over a critically important urban resource (especially amid rapid urbanisation) also makes cities vulnerable to more organised and even more violent armed groups: in Karachi, the Taliban became involved in land grabbing and reselling when they started encroaching on Pashtun peripheries, replicating the involvement of political militias with land grabbers. 38 In Mogadishu, whereas al-Shabaab seems not to be involved in land grabbing or evictions, it benefits politically by becoming a reliable (if feared) adjudicator in the myriad land-related disputes.
Land grabbing not only reflects the unplanned urbanisation patterns exacerbating the scarcity of housing amid massive flows of people into cities but also creates new urban spaces. Both political militias and the Taliban have carved 'slices of land up to create new squatter colonies, then they subsequently sell it off' (Hamid 2015). Graveyards have been 'flattened or flooded with sewage' to pave way for illegal occupation; parks have been converted into new neighbourhood areas; coastal land has been filled with garbage to allow for more informal housing; drainage channels were also used for construction and caused flooding during monsoon season (Hasan et al. 2013: 46). The 'informal developers' conducting this illicit city-making would switch allegiances to the different political parties (and militias) involved in regulating and protecting land grabbing. But, during the intensification of political violence in the city during the period studied (2008)(2009)(2010)(2011)(2012)(2013)(2014)(2015), each developer was 'bound for his own safety and sustenance to one political party…. his lot [was] now tied to theirs' (Hasan et al. 2013: 45).

Political implications of conflict over urban resources
Control over economic resources in cities has come hand in hand with important political advantages. 39 The ability to extract 'taxes' or extortion in exchange for security has provided not only revenue but also political opportunities: for instance, political parties in Karachi were also interested in the electoral advantages of being in control over certain areas. The PPP allied with the PAC, which was a hybrid militia and political group formed by former gang leaders in the area, supporting the PAC with its political influence in Karachi in exchange for access to votes in the neighbourhood (Hussain & Shelley 2016: 12). There, the extraction of bhatta (extortion money), even if exploitative of local populations, was tied to locals' protection against rival political militias and provision of basic services (distribution of rations, medical clinics and support to sports activities) (Kirmani 2017: 110). The business of non-state security provision to low-income areas became so widespread that some areas would face ' competing offers of protection' by different gangs and militias, which in turn led to more violence, intensifying the need for protection while also making the 'service' pro- 35 Interview with senior Somali government official, Mogadishu, 15 October 2019. 36 Interview with non-governmental organisation worker, Nairobi, 19 October 2019. 37 Interview with former national security official, Mogadishu, 16 October 2019. 38 Interview with law enforcement official, Karachi, 09 December 2019. 39 Both groups benefitted financially from the extortion, security taxes and land grabbing activities, though the data does not allow us to lay out exactly how the income from such rents was distributed, whether it went to the central coffers of the armed groups or stayed mostly within their urban-based chapters.
vided by these non-state armed groups less reliable (because no one could fully stop attacks in Karachi's violent context) (Gayer 2012: 80). The ethnic or religious reasons for providing protection to certain groups (for instance, MQM to Mohajirs) became a business as well as a politically motivated practice, as 'the presence of armed men within political or religious groups makes it possible to apply violence to one context (political strife) and then to another (illicit activities) for a limited additional cost' (Gayer 2012: 80). Control over urban resources has also boosted non-state armed groups' political narratives and standing among certain social groups. Al-Shabaab's extortion capacity in Bakara Market, the busiest commercial area in the city and probably the country, benefits its claim of being a more effective (even if brutal) ruler. Al-Shabaab assigns economic and political value in taxing the population: the insurgency's leadership claims 'the government is taking money but is not delivering service to the people'; whereas, '"we, al-Shabaab", take the money from the people and [provide] service to the people'. 40 Whereas al-Shabaab does not provide public services other than dispute resolution in the capital, several interviewees stated that people in Mogadishu regularly resort to al-Shabaab courts and regard them as faster and more effective in enforcing their decisions. 41 At the same time, the absence of effective government regulation over such critical issues as land and housing-and the insecurity caused to those affected-has been in stark contrast to the effectiveness of the insurgents' dispute resolution mechanisms. 42 Another way that people in Mogadishu regard al-Shabaab as more effective is through the aforementioned checkpoints in roads connecting the city to others in Somalia, with government-controlled roads demanding more payments than insurgent-held ones. 43 These parallel regulation and taxation structures were described by one senior member of an international organisation as 'parallel governance structures'. 44 Another political 'benefit' extracted by al-Shabaab from its extortion practices is the heightening of fear linked to terrorist attacks it conducts in Mogadishu. As one senior security official said: 'In order to extort, they have to instil fear and in order to [do that] they slaughter people and put bombs under vehicles of whoever they think they can extort money from'. 45 This fear effect comprises an interesting linkage between economic resources and the political communication role of terrorism, reinforcing a sense (very much present in the minds of people in Mogadishu) that the government is not able to secure vital economic assets and the lives of those linked to them.
Urban resources tied to valuable economic opportunities, therefore, have been simultaneously linked to political benefits. Resources, such as protection rackets, service provision and control over land markets in certain areas, have also become political assets, deriving value beyond money to armed groups controlling them. Votes, support for political clients and stoking anti-government sentiment represented valuable political assets linked to urban areas.
The weakness of the state, in terms of official service provision and institutions (such as dispute resolution), has amplified the political benefits for armed groups to step in. Providing security and some services, as well as resolving disputes related to the critical land issue, has fulfilled acute needs and tied populations of affected areas increasingly to the rules and mechanisms set up by non-state armed groups.

Conclusion
Scarcity of essential materials or services in densely inhabited (but often low-income) areas-water, electricity, physical security, dispute resolution-has been transformed into economic and political opportunities by non-state groups able to wield armed violence. This scarcity does not relate to a natural state of affairs (as in the case of natural resources); it derives to a great extent from unequal and inefficient service provision by the state. Hence the importance of understanding the roles that urban resources play in the behaviours of non-state armed groups and the use of armed violence: the findings here can inform public policies and strategies towards conflict prevention, resolution and socio-economic development in rapidly urbanising settings.
Illicit economies linked to such scarcity have become, in Mogadishu and Karachi, sources of both economic and political benefits for armed groups, with competition for them being a key driver of violence.
In both cities studied, activities often considered ' criminal' due to the income they generate have also produced political advantages in the form of insurgent propaganda in Mogadishu, ethnic or clan-based spatial claims in Karachi and Mogadishu and votes for political militias in Karachi.
Urban resources underpin a particular sub-set of conflict economies tied to spatial and economic features unique to large cities. They perform an important function for armed groups: they provide economic incentives for violent activity and territorial presence while linking these armed actors to the politics of the urban environment. Because these resources are tied to social, economic and political traits specific to urban areas, especially scarce public services amid rising populations, they comprise a particular aspect of the political economy of conflict that deserves more attention from policymakers and experts alike.
The concepts of urban resources and cities' functions for armed groups are aimed at enriching discussions about the unique challenges of urban environments for conflict and post-conflict settings. This article has provided further evidence of the specificities of armed violence in cities by showing how economic and political activities have reinforced each other in armed groups' urban strategies.

Funding Information
This article is based on primary data (interviews) collected during field research as part of a project funded with aid from the UK government; the views expressed do not necessarily reflect the UK government's official policies. The project, titled 'Urban Drivers of Political Violence' has been conducted through the International Institute for Strategic Studies.