This book has explored subnational variation in party violence in Pakistan through focused case studies of four major Pakistani political parties: the Muttahida Qaumi Movement (MQM), the Pakistan Peoples Party (PPP), the Pakistan Muslim League-Nawaz (PML-N), and the Awami National Party (ANP). These parties were highlighted because they each showcased a different party violence strategy, with two of the parties (the PPP and ANP) exhibiting subnational variation in such strategies associated with the varying benefits and costs of violence in different parts of Pakistan. Yet, these four parties hardly constitute the universe of potential cases in Pakistan, let alone globally. A major motivator of this research has been the sad reality that political parties engage in violence in numerous countries around the world. To fully assess the local-level power dynamics that lead to such outcomes in other states would require extensive multi-method research in each country – efforts that are beyond the scope of this project. Nonetheless, it is useful to examine how specific aspects of the argument presented in this book may be applied to explain electoral and party violence more broadly than the case studies offered thus far, including by examining party violence in other developing democracies. In doing so, I am able to demonstrate the extent to which the theory’s explanatory power helps us understand cases that are intrinsically important and also show that the theory is generalizable to contexts beyond the parties studied herein.
My goals in this chapter are accordingly limited to demonstrating and bounding the external validity of my argument. To do so, first, I examine how my theory travels to a Pakistani case not covered so far, that of the Pakistan Tehreek-e-Insaaf (PTI), a party that was beginning to gain influence in the years I was carrying out my research and that won national elections in 2018. Then, I focus in more detail on the strategies of political parties in three multiethnic and multiparty states – Nigeria, the Philippines, and India – and offer them as illustrations of the potential gains of studying political processes through the approach offered in this book.
Nigeria and the Philippines are important countries to study the determinants of party violence. Both states lack complete monopoly over the legitimate use of violence within their borders and their societies are characterized by numerous, politically salient social divisions. The countries have also suffered from high levels of electoral and party violence in recent decades, manifesting in both electoral alliance formation and violence outsourcing. Yet, party politics also differ in important ways. The role of political parties in the Philippines, in particular, is considered secondary to local elite politics, with one scholar describing parties as virtually nonexistent (Kenny Reference Kenny, Carothers and O’Donohue2020). Nonetheless, we see patterns of alliance formation between parties and elite violence specialists, which can be explained using a similar logic to that observed in Pakistan.
The countries are also intrinsically important. With over 200 million people, Nigeria is the most populous country in Africa and the seventh most populous in the world. Like Pakistan, Nigeria is embroiled in a number of different conflicts: from armed conflict with the militant Boko Haram, to local violence between nomadic Fulani herdsmen and indigenous farming communities, to criminal violence in urban centers. Political parties thus choose strategies against this violent backdrop. Similarly, the Philippines, an archipelago in Southeast Asia, has faced a violent insurgency in the Muslim-majority Mindanao region of the country, substantial violence associated with organized crime, as well as state-orchestrated violence. Landed elites have also long wielded violence at the local level, a trend that intersects with electoral politics in important ways.
Finally, in India – comparatively a more established democracy – I focus on the Shiv Sena, an organizationally strong party that maintains a captive support base among Marathi-speaking voters in Mumbai, and accordingly, engages in direct violence in this urban center. However, as it has expanded to other parts of Maharashtra state where its “sons of the soil” appeal holds less resonance and where, to a large extent, it is unable to replicate its organizational prowess, it has adopted different strategies, consistent with the logic of my theory. The Shiv Sena is particularly useful because it has invited comparisons to the MQM, a party that is often regarded as exceptional both in its use of violence and in its organizational reach.
The sections that follow should be read as “shadow” cases rather than as full empirical tests of my theory. In each case, I use secondary literature and/or survey data to argue for the presence of the various violence strategies articulated in this book and to show that party organizational structure, nature of party support base, and political landscape are instrumental in determining whether a party engages in violence and the strategy of violence that it pursues. The full complexity of the relationship between political parties and violence in each of these countries undoubtedly deserves more treatment than I can offer here. Yet, the evidence I present is sufficient to demonstrate that the logic of political party violence observed in the Pakistani experience is frequently found elsewhere as well.
8.1 The Out-Of-Sample Pakistan Case: the Pakistan Tehreek-E-Insaaf
This book has focused primarily on four political parties operating in different parts of Pakistan. In this section, I explore the rise to prominence of a relatively new political party, the PTI – a rise that happened contemporaneously with my research.
The PTI was formed in 1996 by Imran Khan, a former national cricket player who famously led Pakistan’s national team to victory in the 1992 Cricket World Cup. Despite Khan’s personal appeal and charisma, in its early years, the PTI had just a small cohort of activists. It performed poorly in the 1997 and 2002 elections, during which time the party won only Khan’s seat in Mianwali in Punjab. Between 2002 and 2011, the party attempted to build a grassroots level organization aimed at mobilizing its supporters. However, soon thereafter, it began to attract the entry of professional politicians, or electables, from other mainstream parties, most notably the PPP, the PML-N, and the party formerly affiliated with Gen. Pervez Musharraf, the Pakistan Muslim League-Quaid (PML-Q). Tabinda Khan (Reference Khan, Mufti, Shafqat and Siddiqui2020, 69) writes that the entry of these electables “strained and ultimately unraveled the PTI’s nascent internal democracy and fundraising structure and reduced the momentum of its membership campaign.” A combination of their inclusion in the party and efforts made by Imran Khan to consolidate his power within the party structure led PTI party workers “to leave the PTI’s grassroots structure in droves” (Khan Reference Khan, Mufti, Shafqat and Siddiqui2020, 69). In 2013, the party received the third highest number of seats in the National Assembly, effectively displaced the PPP in Punjab as the PML-N’s most serious opposition in that province, and formed the government in Khyber Pakhtunkhwa (KP).
After this success, Khan grew “convinced that the strategies of traditional political parties – including the recruitment of moneyed electables, reliance on large donors, and centralized top-down decision-making and election management – were necessary to win elections and he put the party’s institutionalization on the back burner” (Khan Reference Khan, Mufti, Shafqat and Siddiqui2020, 70). The party therefore ended up relying heavily on landed and business elites to form and maintain clientelistic linkages with voters at the local level, rather than investing further resources into creating a local party apparatus. Many PTI leaders and prominent members have come under fire for criticizing the PML-N and PPP for representing the “old elite” given that they themselves once belonged to those parties. Of the 92 PTI members of National and provincial assembly that I surveyed in 2020, 52 percent said that they had previously been affiliated with other parties in the past, primarily the PML-N, PPP, and PML-Q. The party also lacks party offices and local presence in many districts in which it competes, and relies heavily on the candidate’s own finances to run elections. As such, the party can be characterized as weakly organized.
The PTI does not rely on the votes or support of only certain identity groups, choosing not to actively exclude any voters in the electoral arenas in which it operates. I have shown in Chapter 4 how, even in Karachi – where most parties form ethnic linkages – the PTI primarily led a multiethnic campaign appealing to members of all different ethnic groups. News sources, for example, referred to the PTI’s success in 2018 as evidence that this was a “post-ethnic election” (Salam Reference Salam2018). According to a survey conducted in Karachi in the days leading up to the 2018 elections, sizeable percentages of each ethnic group said that they would vote for the PTI (see Figure 8.1). By early 2021, however, focus groups in Karachi found that some voters, particularly Muhajir voters, who had voted for the PTI in the 2018 election were disillusioned by the party’s governance and were now looking to switch their vote elsewhere, including back to the MQM.
Figure 8.1 Distribution of ethnic support for the PTI in Karachi
Given its weak organizational structure and this absence of a captive support base, the PTI is unlikely to engage in violence directly or by outsourcing to violence specialists – even in Karachi where other parties with captive support bases have regularly done so. Prior to the 2018 elections, the PTI did indeed refrain from ethnic or electoral violence, leading a largely peaceful campaign. Overall, party violence in Karachi was much lower than it had been in previous years and election cycles. This, no doubt, had much to do with the fact that the military had cracked down upon a violent MQM and People’s Aman Committee (PAC), and that the ANP had effectively been wiped out by the Pakistan Taliban, disrupting the ethnic nature of electoral politics that had long characterized the city. Nonetheless, that the PTI did not enter into electoral politics or succeed electorally in Karachi by using violence indicates that the explanation for party violence lies at least partly with the nature of parties’ relationship to voters rather than in factors exclusively related to Karachi city.
While the PTI has refrained from direct or outsourced violence in Karachi, the PTI’s weak organizational structure means that it has the incentives to form alliances with local-level patrons, including elite violence specialists, where these actors control vote banks. In the weeks and months preceding the 2018 elections, the PTI did reach out and form electoral alliances with members of the sectarian Ahle Sunnat Wal Jamaat, which, as explained in Chapter 6, served as a valuable ally to the PML-N during the 2008 and 2013 elections. An ASWJ spokesman claimed that the group had supported the PTI in no less than 70 constituencies in 2018 and also boasted that “most of its supported PTI candidates won their seats on July 25” (Altaf Reference Altaf2018). Numerous pictures of meetings between PTI candidates and local ASWJ leaders surfaced just prior to the national election and subsequent by-elections.Footnote 1 For example, the official social media handle for the ASWJ tweeted pictures of the PTI candidate for a by-election in Islamabad meeting with a senior leader of the ASWJ. A news report noted, “According to the ASWJ’s spokesperson in Islamabad, the group has not yet reached a conclusion on giving support to any candidate and that the decision would be made after consultation with party leadership” (Pakistan Today 2018). Another PTI candidate went so far as to publish a newspaper advertisement in which he thanked the ASWJ for supporting him during a by-election in Haripur in KP in 2015. Even after coming to power, the Minister of State for Interior was seen sharing the stage with Ludhianvi, head of the ASWJ, just a day after a terrorist attack on the Shia Hazara community in Quetta. Not unlike the PPP and PML-N who have also engaged in electoral alliances with these militant actors, the PTI denies allegations of any agreements with the ASWJ. The party stated that it had sent a “show cause” notice to one of its MPA candidates for seeking the support of the ASWJ. However, it nonetheless admitted that while “it’s not a party position … there can be discussions at individual level” (Altaf Reference Altaf2018b).
The PTI’s engagement in violent strategies provides further evidence that existing political conditions coupled with a party’s internal capacity and support base structure the incentives for parties to engage in violence. As the nature of the party’s relationship with its voters continues to evolve – and especially if the polarizing nature of its politics results in a captive support base – we should expect to see its tactics change accordingly. Examining this party thus allows us to assess the future direction of party violence in Pakistan, a task I take on further in Chapter 9. I now move to examining party violence in contexts outside of Pakistan.
8.2 Party Violence in Nigeria
Nigeria is a federal republic with a multiparty system. From 1966 until 1999, Nigeria remained primarily under military rule. In 1993, nine years after a coup overthrew Nigeria’s Second Republic, multiparty elections were held in the country. This democratic experiment was exceptionally short-lived – lasting only months – as the military took over again in a “brief but extraordinarily predatory period of military rule” (LeBas Reference LeBas2013, 251). When multiparty democracy returned to Nigeria again in 1999, it was accompanied by high levels of electoral and party violence which have continued till this day.
The country’s history of military coups has led, as in Pakistan, to a powerful military that still maintains considerable influence even during ostensibly democratic set-ups. Nigeria is generally considered a hybrid regime or one of competitive authoritarianism. Elections to the House of Representatives and Senate take place in first-past-the-post voting in single-member constituencies. Trust in the armed forces is higher than trust in either ruling or opposition political parties, although about 47 percent of respondents from the 2018 Afrobarometer survey said that they felt close to a particular political party. Political parties therefore operate within these confines and face institutional uncertainty and unclear time horizons (Lupu & Riedl Reference Lupu and Riedl2013), which often leads to underinvestment in party organizational strength. Additionally, violence specialists have arisen in the gap left by an absent, or uninterested, state, resulting in a lack of monopoly over the legitimate use of violence and political landscapes of shared sovereignty. The brutal nature of General Sani Abacha’s regime from 1993 to 1998 “led ordinary Nigerians to look for security and political representation via the formation of ethnic militia” (LeBas Reference LeBas2013, 251).
With over 250 ethnic groups, Nigeria is a society divided along numerous ethnic cleavages. Two-thirds of its approximately 200 million people belong to one of three ethnicities – the Hausa, Yoruba, and Igbo. Survey data shows that Nigerians are more likely to say that they identify with their ethnic identity than the national identity. A slight majority (52%) believe that their ethnic group is sometimes, often, or always treated unfairly by the government, and about 28 percent believe that they were personally discriminated against on the basis of their ethnicity in the past year (2018 Afrobarometer). Many parties have ethnic and regional support bases. Few parties are differentiated along ideological or programmatic lines (Turnbull Reference Turnbull2020). Religion is another salient identity, with the country nearly evenly divided between Christians and Muslims, with the former tending to predominate in the south and the latter in the north.Footnote 2 Finally, violent conflict between nomadic herders and sedentary agrarian communities is also common. Which identity cleavage is locally salient varies around the country. This section cannot do justice to the range of identities in Nigeria, nor to the wide variety of violence specialists operating in different parts of the country. The ethnic militias discussed here are primarily chosen for illustrate purposes.
There is much evidence to suggest that political parties in the country engage in violence before and during elections but also the periods in between. Violence takes the form of political killings, armed clashes between supporters of opposition parties, and even bombings. Ethnic and religious riots are also common in many parts of the country. Violence is so commonplace, in fact, that politicians discuss their involvement in it openly – one opposition candidate told Human Rights Watch, “That’s the way it works here … If anyone tries to attack me, my boys will unleash terror” (Human Rights Watch 2007a). In 2018, nearly 34 percent of Afrobarometer respondents feared “somewhat” or “a lot” becoming a victim of political intimidation or violence during election campaigns in the country. A majority of respondents (52.5%) also believed that the government was handling the prevention of political violence during election campaigns either “very” badly or “fairly” badly. Polling-day violence is common; polling agents have even been killed by thugs hired by politicians. The police appear to both lack the capacity and the willingness to crack down upon violent actors. A police official was quoted as stating that the problem of political violence was “more or less of politics, not one of law and order” (Human Rights Watch 2007a). Intraparty violence also occurs frequently, with fierce – and often violent – competition taking place over nominations to the ruling party’s seats in national and state assemblies (Cohen Reference Cohen2016).
A 2007 Human Rights Watch report found that “many political figures openly recruit and arm criminal gangs to unleash terror upon their opponents and ordinary members of the public” and that “in some states, powerful and violent political ‘godfathers’ have gained control over politicians who are dependent on those sponsors to provide protection and fight their street battles” (Human Rights Watch 2007b). Both of these strategies of violence – outsourcing to ethnic militias and alliance formation with elite violence specialists – are familiar to patterns we see in Pakistan. Sections 8.2.1 and 8.2.2 will provide an overview of these strategies in Nigeria by focusing on some prominent political parties and the violence specialists on whom they have relied. While both the outcomes and determinants of party violence have significant degrees of overlap with Pakistan, there are also some key differences that I outline in the discussion that follows.
8.2.1 Party Outsourcing and Violence Specialists in Nigeria
The 2019 elections in Nigeria were accompanied by significant violence, consistent with levels of violence seen in previous elections. Just prior to them, the National Security Adviser urged politicians to refrain from “mobilizing merchants of violence, including armed bandits and terrorists to orchestrate violence during the elections in several states of the federation” (Mutum & Okar Reference Mutum and Okar2019). LeBas captured this phenomenon of party outsourcing when she wrote a few years earlier, “For many politicians … militias represent an attractive political resource, and there have been consistent reports since 1999 of the use of the [Oodua People’s Congress] and other militia to intimidate voters, organize poll fraud, and threaten rival candidates” (2013, 253). Who are these violence specialists upon whom parties and the state rely during election time, and why do they do so? This section turns to these questions next by examining a few examples of political parties and violence specialists.
8.2.1.1 The People’s Democratic Party and Violence Outsourcing
The People’s Democratic Party (PDP), founded near the end of military rule in 1998, was until recently the most prominent political party of the Fourth Republic, winning every Presidential election between 1999 and 2011. It was also the governing party until 2015 and is generally considered center-right.
The PDP is a classic example of a weakly organized party. Aleyomi (Reference Aleyomi2013, 290) writes that the party has no clear ideology, is afflicted with intraparty competition and conflict, lacks a transparent selection process for candidates and party leadership, and “has been inadvertently hijacked by plutocrats and kleptocrats.” As a result of this weak structure, it relies on “prominent and wealthy individuals with strong patron–client networks who facilitate the election of their associates to high office” (Kendhammer Reference Kendhammer2010, 62). Party switching between the PDP and opposition parties, such as the All Nigeria Peoples’ Party (which merged with other parties to form the All Progressives Congress in 2013), is common, and socialization procedures of party members are largely nonexistent. The lack of local-level party presence and absence of socialization procedures means that the party does not have the ability to operate clientelistic networks, mobilize voters, or directly control agents of violence.
While the PDP purports to be a mainstream, multiethnic political party – its ruling coalition incorporates elites from the largest ethnic groups in the country – in key constituencies around the country it nonetheless relies on voters belonging predominantly to one ethnic community. In many electoral arenas, PDP candidates engage in ethnic “talk” by running “campaigns based on mobilising ethno-religious identity groups and … their supporters frequently participate in ethnoreligious violence” (Kendhammer Reference Kendhammer2010, 50). For example, while the PDP had a Muslim Hausa president, in the ethnically and religiously diverse city of Jos, the capital of Plateau State in central Nigeria, it was viewed as the main party representing the indigenous Christian population. Meanwhile, in Zamfara, the PDP was seen as a Hausa-Muslim party and its affiliated politicians and party workers used “attacks on Christians to curry favour with their constituencies” (Kendhammer Reference Kendhammer2010, 49). This created seemingly contradictory national policies, “cooperative at the federal centre, but ethnically antagonistic at home” (51). As with the PPP in Pakistan, then, the PDP could count on a relatively captive support base composed of one type of ethnic voter in one arena but this did not travel to all arenas in which it contested.
The PDP faced varying levels of cost for its use of violence across different electoral arenas. Kendhammer (Reference Kendhammer2010, 67) writes that the PDP had “little reason to intervene decisively against the corruption, violence, or ethnic chauvisnism of its ‘representatives’ in the states, especially those with substantial strategic importance for the PDP’s electoral ambitions” in part because violence was “not political destabilizing for the PDP coalition because the benefits of political mobilization resulting from the violence are not outweighed by the need to incorporate these local minorities into the party.” However, despite its incentives to permit and facilitate ethnic violence in certain arenas, Kendhammer (Reference Kendhammer2010) argues that the PDP was forced to “compartmentalize” the exclusionary rhetoric and violent actions taken by some local elites on behalf of the party in order to maintain its reputation of being a multiethnic party. This required PDP politicians to be cautious in their overt support for ethnic “talk” and violence in local arenas to ensure that they did not upset elites that were part of their multiethnic coalition at the federal level. Kendhammer writes:
Where this shield is weakest – where its politicians become most obviously tied up in the business of ethnic organisations and militias (the South-East) – the PDP has suffered electoral setbacks.… By walling off most (but not all) ethnic ‘talk’ from the PDP itself, it is far easier for the larger ethnic factions in the PDP – most importantly the Muslim Hausas – to demonstrate to minority partners that the coalition is not merely a ploy for them to assert their dominance.
These dynamics are evocative of this book’s argument in a number of ways. Ethnic violence in parts of Nigeria can result in political mobilization of co-ethnic voters, resulting in party vote gain and limited voter costs. The party is, thus, able to uphold a “multiethnic” identity at the national level even while it purports to represent only certain groups of voters in some states or regions. However, just because the PDP benefits from this violence in the arena in which violence is being enacted doesn’t mean that it doesn’t face reputational costs from outside that arena. As such, the party follows distinct strategies of violence in different arenas and must also maintain distance from any ethnic violence in which it engages for purposes of plausible deniability.
Because of the PDP’s weak organizational structure and this need to maintain distance from the act of violence, the party relies on violent specialists to engage in violence on its behalf. These specialists vary depending on the region in which the PDP is operating. I focus briefly on the predominantly Igbo Bakassi Boys, upon whom the PDP and many other parties have relied in the southeastern part of the country. Formed in 1998 in Abia State, the Bakassi Boys initially had the narrow aim of curtailing extortion and armed robbery at local markets (Smith Reference Smith2004). Soon, however, they expanded their influence into nearby provinces, such as Anambra, and began to receive support from the state government in the forms of funding and weapons. In return, the militia was expected to target the governor’s political opponents: “the governor … allegedly used the group to hunt down his political rivals and critics and provided protection for members who used their positions to loot and abuse citizens” (Reno Reference Reno2011, 207).
The Bakassi Boys were initially very popular among the populace, particularly the co-ethnic Igbo community, as they were perceived as tackling crime and responding to local demands in a way that politicians and the police had failed to do. They were also known for using magic and superstition, which led to their perception as “superheroes” (Sandve Reference Sandve2009, 49; Smith Reference Smith2004). Like the PAC in Karachi, then, the Bakassi Boys had a Robin Hood-esque persona whose presence restored security in crime-ridden areas. Their support was in stark contrast to that of elected legislators who had little local influence among the populace. A legislative head of a political party explained that the “majority of us cannot go home. You drive your NASS (National Assembly) car on the streets (and people) shout ‘thief, thief’” (Onyeacholem, quoted in Reno Reference Reno2011, 208). This anecdote evokes that of PPP legislators being unable to enter their own constituencies in Lyari.
Political parties frequently sought the support of the Bakassi Boys in an effort to increase their own local standing. Indeed, “all sorts of politicians and businessmen make use of the Bakassi Boys or other armed gangs” (Harnischfeger Reference Harnischfeger2003, 27). In some cases, the vigilante group was used to kill members of other ethnic groups. By some accounts, the Bakassi Boys had executed a startling 3,000 individuals in Anambra by 2003 (Harnischfeger Reference Harnischfeger2003). In other cases, politicians turned away from local patrons who had previously existed in the area and sought to cooperate instead with the Bakassi. For example, Governor Mbandinuju of the then-ruling party PDP was initially reluctant to allow the Bakassi Boys to operate. He had come to power in part due to support from another local patron, Emeka Offor, a rich businessman. In exchange for Offor’s support, the governor promised to allow him to appoint two ministers in the future cabinet. After the election, the governor reneged on his promise, and turned to the Bakassi Boys to target and kill his political opponents.
However, as the use of Bakassi Boys’ violence intensified, local support for them started to dip. Like the PAC, there were often power struggles between the gang and the parties who outsourced violence to them. That the Bakassi Boys work with numerous different political parties – all appealing to different ethnic groups – further indicates that the relationship between the party and the militia was not maintained only as a result of ideological or ascriptive reasons.
8.2.1.2 The O’Odua People’s Congress
The O’odua People’s Congress (OPC) is an ethnic militia that was formed in 1994 to represent the interests of the Yoruba, who constitute approximately 15–20 percent of the country’s population and are concentrated in Nigeria’s southwest. By the time that multiparty democracy returned to Nigeria in 1999, the OPC had adopted vigilante and security roles, in some areas taking on responsibilities normally held by the state, such as offering security and arbitrating legal disputes. Local branches of the OPC were spread across southwestern Nigeria, and it operated deep social networks. Primarily composed of unskilled workers and present and former members of gangs in Lagos (Adebanwi Reference Adebanwi2005), some estimate that it had more than 3 million members at its peak (Akinyele Reference Akinyele2001, 626). In 1999, it gained notoriety and public attention when it engaged in large-scale ethnic clashes with Hausas, backed by another militia, the Arewa People’s Congress. Sandve writes, “the ethnic militia had gained much credit for containing violence albeit by perpetrating its own violence” (2009, 43).
The OPC has functioned as an archetypical street-level violence specialist to whom weakly organized ethnic political parties have outsourced the tasks of intimidating voters, threatening rival candidates, and ballot stuffing. On occasion, the OPC and rival ethnic militia have played key roles in riots and “large-scale ethnic confrontations” (LeBas Reference LeBas2013, 253). Reno writes that the OPC “acted more as a tool of politicians’ ambitions rather than as a new force in politics” (2011, 206) and while its members “organized in response to community grievances against the state, their actions often served the interests of a politician or clique” (207). Similarly, Guichaoua (Reference Guichaoua2010) explains, on the basis of interviews with OPC members, that “several of our OPC informants mentioned the actions they personally took, against payment, to coercively induce voters to opt for their local political patron during the 2003 general elections.” Politicians have also been able to benefit by taking credit for the services that these militias provide to the public. The OPC, meanwhile, has been forced to balance its image as a supporter and representative of the local community with its desire to receive patronage and access to the state through collaboration with state actors.
How do voters respond to the use of violence by the OPC and other violence specialists, as well as party patronage of these actors? Afrobarometer survey data on levels of acceptability of violence analyzed in LeBas (Reference LeBas2013) demonstrates that there are stark differences between views in urban slums where these groups are influential and in other, more rural areas. While rejection of violence is “overwhelming” in Nigeria, residents of slums in the megacity of Lagos “do not seem equally unified in their support of this societal norm” (257). These results are consistent with my findings in Karachi in relation to the rest of Pakistan. While LeBas views this discrepancy as indicative of the complex relationship between security provision and violence specialists in this “relatively ungoverned” urban space (240), which “allows greater latitude for local armed actors to play security or enforcement roles” (257), rather than of the role played by ethnic cleavages, the militias nonetheless capitalize on ethnic solidarities among voters, consistent with this book’s arguments.
8.2.2 Alliance Formation with Elite Violence Specialists
In addition to parties outsourcing violence, we also see examples of parties forming alliances with elite violent specialists in Nigeria. Nigerian politics is characterized by the presence of what are known as godfathers – local elites who have “respect and follower (voters) in the community” and without whose support “very few politicians can achieve success” (Majekodunmi & Awosika Reference Majekodunmi and Awosika2013, 70). Historically, godfathers provided financial support and lent their backing to electoral candidates (their “godsons”) and also took on the role of mobilizing voters. In return, godsons shared public loot, awarded contract to their godfathers, and enabled political appointments (Majekodunmi & Awosika Reference Majekodunmi and Awosika2013, 71). This quid-pro-quo relationship was intended to be mutually beneficial to the godfather and the political party. In more recent times, especially since 1999, the godfather has been able to use his power to move away from being a mere “broker” to capturing the “party machinery for himself” by becoming the sponsor of individual politicians (Olarinmoye Reference Olarinmoye2008, 70). As such, godfathers are able to decide who gets nominated to contest elections as well as who wins the election. The dynamic between godfathers and political parties has become more unequal as a result. A BBC report from 2019 wrote about one godfather that “on many campaign posters, Mr Kwankwaso’s image is larger than that of the PDP contender” (Jones Reference Jones2019; see also Ibrahim Reference Ibrahim2003).
Godfathers have local influence and wealth. Their influence stems, Turnbull (Reference Turnbull2020, 46) writes, “from their immense personal wealth, extensive patronage networks, and connections to the national party leadership.” Because most political parties have weak organizational structures, they are unable to operate clientelistic structures or rally supporters themselves. One scholar attributes the relationship between party candidates and godfathers to this lack of party capacity; Olarinmoye (Reference Olarinmoye2008, 68) writes that parties have “reduced mobilization capacity as they are poorly entrenched in Nigeria civil society with a gap existing between them and the masses forming the electorate.” Because maintaining patron–client relations “demand[s] institutional and human resources that are beyond the capacity of African political parties as they are poorly institutionalized,” parties have resorted to allying with godfathers “to effectively deploy patronage and overcome the problems of voters’ compliance, which it generates” (Olarinmoye Reference Olarinmoye2008, 67–69). Godfathers have an “intuitive grasp of the structure of the voting public in the sense of his knowing what appeals and what incentives are most appropriate to each group” and the “capacity to deploy such knowledge to ensure voter compliance” (Olarinmoye Reference Olarinmoye2008, 69). Thus, consistent with this book’s theory, alliances with patrons are necessary for organizationally thin political parties to mobilize voters, manage clientelistic exchange, and engage in constituency service.
Like some feudal elites and sectarian militants in Pakistan, godfathers in Nigeria utilize violence as one means to achieve control and can be considered violence specialists. Human Rights Watch writes, for example, that “Oyo State political godfather Lamidi Adedibu has retained his position of political power within the PDP and Oyo politics as a whole despite his open involvement in fomenting political violence in the state” (2007, 43). Indeed, their “power stems not just from wealth but from their ability to deploy violence and corruption to manipulate national, state or local political systems in support of the politicians they sponsor” (Human Rights Watch 2007b, 33). Godfathers engage in “violent scheming and aggressive ‘politiciking’, coupled with manipulating devices of having their way by any means” (Majekodunmi & Awosika Reference Majekodunmi and Awosika2013, 71).
Where the godfathers themselves employ street-level violence specialists to carry out violence on their behalf, the dynamics of violence outsourcing and alliance formation overlap. Indeed, it is not unusual for godfathers to rely on distinct militias to engage in the dirty act of violence. Here, the relationship between the party and the godfather can still be characterized as one of alliance formation for purposes of vote gain, but the godfather may rely on a street-level violence specialist to perpetrate violence on his behalf. In Pakistan, too, we observe feudals working with dacoits as one means to maintain their localized political and economic hegemony against rivals. Turnbull (Reference Turnbull2020) explains that the informal delegation of violence allows godfathers themselves to avoid accountability among their clients.
8.2.3 Summary
Despite important differences, the Nigeria case shows support for several observable implications of my theory. Political violence in Nigeria is pervasive. Parties engage in both alliances with elite violence specialists and outsource violence to street-level violence specialists. In some cases, these two modes of violence overlap: Parties ally with godfathers, who in turn outsource the task to other local actors. Scholars have provided different explanations for these phenomena. Turnbull (Reference Turnbull2020) explains that intraparty competition determines which gubernatorial elections are highly competitive and in turn, when parties rely on “social movement leaders” to mobilize voters and orchestrate pre-election violence. Without discounting these accounts, the Pakistan case encourages us to focus on the role of parties’ weak organizational structures in explaining party reliance on godfathers who control local vote banks and patronage networks. While more complex, the role of ethnic “talk” and captive support bases also point to how violence can be varyingly beneficial for parties, often useful in the local electoral arena in which violence is being utilized but costlier at the national level where the party’s image as a multiethnic party is more relevant. Finally, the logic of reputational costs of violence pervades many explanations of violence outsourcing in the country, though here, as Kendhammer (Reference Kendhammer2010) argues, the primary cost calculus is one which takes into consideration costs imposed on the party by ethnic party elites rather than voters. By creating distance between the ethnic rhetoric and the ethnic violence – in part by outsourcing violence to distinct violence specialists – the PDP signals to its coalition partners that it takes its multiethnic identity seriously. Turnbull (Reference Turnbull2020) similarly explains the process of outsourcing partly through this lens when she writes that “non-state actors are more appealing partners because they offer elites plausible deniability, helping [parties] mitigate the reputational costs of repression.”
8.3 Party Violence in the Philippines
Like Nigeria and Pakistan, the history of the Philippines is peppered with military coups. In 1972, President Marcos announced martial law in the Philippines, bringing an end to the country’s experiment with multiparty democracy that had begun in 1946 following independence from US control. This authoritarian interlude lasted until 1986, when the People Power Revolution brought hope that the restoration of democratic institutions would herald a new era of politics. The 1987 constitution established a multiparty system under a presidential form of government. Today, separate elections are held for the president, vice president, and senators, each elected for six-year terms, with members of the House of Representatives, governors, and other local-level officials elected to serve three-year terms. Elected officials are chosen by a direct vote through a first-past-the-post system.
The last 30 years have also been marked by high levels of electoral and party violence. In particular, provincial and local elections – which bring to power provincial governors, city and municipal mayors, and individuals on municipal and city council seats – have seen large-scale violence. Villanueva writes, “An election without violence and casualties would be a miracle in Philippine political campaigns” (1996, 182). Political violence in the Philippines takes many forms (Patino & Velasco Reference Patino and Velasco2004). Political rivalries routinely result in gunfights during election campaigning, often “related to the intensely personal characteristic of Philippine politics, especially at the local level” (Rocamora Reference Rocamora and Lacaba1995, XIV). Opposition supporters are threatened and intimidated. Politicians engage in kidnapping for ransom in order to raise funds. Communist parties have armed wings that they use to wage war against ideological opponents. Politicians are even known to have their own private armies (Onishi Reference Onishi2010). Kreuzer (Reference Kreuzer2009) notes that between 100 and 200 lives have been lost in each election, with 10–30 percent of these killed being candidates or office-holders. The lead up to the 2009 election resulted in so much violence that the government declared a state of emergency in two southern provinces in the country. Why do we observe such levels of violence? And given the relative unimportance of political parties and party labels in the country, what form does such violence take? In what follows, I provide examples of parties in the Philippines forming electoral alliances with elite violence specialists and explain how multiparty elections have served to strengthen the influence of these actors rather than supplant them.
8.3.1 Party Alliances with Violence Specialists
Scholars and observers of Philippine politics are unanimous in their description of political parties as weakly organized. Manacsa and Tan (Reference Manacsa and Tan2005) pinpoint the “weak internal organizations” of the Philippines’ political parties as one of the reasons for the overall weakness of the country’s party system. Parties are effectively indistinguishable on ideological or programmatic grounds; they do not have carefully defined platforms, often entering into alliances with other parties and forming coalitions. Party-switching (locally known as turn-coatism) is common, with the majority of members of mainstream political parties “not loyal to any party … seizing any opportunity to join the party in power” (Ana III Reference Ana III2019). Parties must also find their own monetary sources, which requires them to embrace candidates who can fund their own campaigns or seek backing from rich businessmen.
Political parties in the Philippines are then merely the personal vehicles of political dynasties: “candidates are not elected because of what their party stands for. The electorate, in general, wants someone on whom they can lean to address their material concerns and needs” (Leones & Moraleda Reference Leones, Moraleda, Sach senrôder and Frings1998, 291). In the absence of well-institutionalized political parties with local-level presence, local elites have become necessary allies to political parties seeking to gain power. National-level candidates “depend on the vote-mobilizing capacity of allies at the local level, since they cannot rely on political parties to do so” (Patino and Velasco Reference Patino and Velasco2004). Both historically and today, local patron–client dynamics contribute to frequent party-switching, “for when a political patron found it advantageous to move from one party to the other, he could count on his clients to follow him” (Landé Reference Landé2002, 121; see also Landé Reference Landé1965).
The primary patrons and violence specialists in the Philippines are feudal elites who have long dominated the country and own large swaths of agricultural land. Sidel describes them as “predatory power brokers who achieve monopolistic control over both coercive and economic resources within given territorial jurisdictions or bailiwicks” (1999, 19). Colonial policies served to strengthen the power of these local elites at the expense of a well-functioning bureaucracy, resulting in a weak central state, with local centers of powers dispersed throughout. American colonial officials writing in 1947 noted that local police forces had long acted as the private forces for the feudals: “Underpaid, only partially trained, and poorly equipped […] they often served the reigning local cacique rather than the public. In numerous instances they were agents of oppression rather than agents of land” (quoted in Kreuzer Reference Kreuzer2009, 51). While there is disagreement among scholars of the Philippines about the exact source of the influence of these strongmen – and the extent to which it derives today from land or more recent institutional choices – it is clear that the state, as Kreuzer (Reference Kreuzer2009) writes, “never enjoyed a monopoly on the means of violence. Alongside legal state agents of coercion there were always private and hybrid organizations that employed violence for a variety of political and economic aims” (48). These violence specialists – which Kreuzer (Reference Kreuzer2009) refers to as private armed groups (PAGs) – have long operated in Javier Auyero’s “grey zone of politics” (2007), both relying on and needed by politicians and the state.
In the postcolonial era, landed elites have further entrenched themselves in the local political economy, often through illegal activities or acts such as heavily regulating crops and industry, among others (Sidel Reference Sidel, Harriss, Stokke and Tornquist2004). As a result, they enjoy both substantial political and economic influence and continue to dominate patron–client relations at the local level (Leones and Moraleda Reference Leones, Moraleda, Sach senrôder and Frings1998, 291). These patron–client relations are “characterized less by mutual benefits and more by coercive force” (Manacsa & Tan Reference Manacsa and Tan2005, 749). Many elite families have armed men under their control. Because the landed elite maintain their own coercive apparatus, the police are largely subservient to them. Many landed elites are related to one another, ensuring that politics in the country is controlled by a handful of dynastic families. Mendoza et al. (Reference Mendoza, Beja Jr., Venida and Yap2016) find that “representatives from political dynasties account for 70 percent of the jurisdiction-based legislators in Congress.” Because the gains from winning office are high, long-standing feuds between these oligarchic clans are common. As a result, these actors increasingly seek to maintain control over elected office, which allows access to state resources.
Multiparty elections have served to strengthen rather than supplant the power of these local elites. The political system that President Corazon Aquino (1986–1992) brought with her after 20 years of military rule “continued to discourage the emergence of stronger, more programmatic political parties” and “facilitated the restoration of the power of local clans … the new parties that did emerge remained remarkably similar in their orientation toward patronage, reliance on coalitions of local elites, nonideological character, and shifting membership” (Hutchcroft & Rocamora Reference Hutchcroft and Rocamora2003, 278). Electoral politics merely “provided the landed elite with a venue to consolidate and expand their power” (Teehankee Reference Teehankee, Croissant and John2002, 155) and, as such, “violence and intimidation, vote-buying and electoral fraud work in tandem with the mobilisation of local machines for self-perpetuation in office” (Sidel Reference Sidel, Harriss, Stokke and Tornquist2004).
More so than Pakistan, where party labels certainly matter to some extent – even if party switching is rampant and organizational structures weak – personalities trump partisanship routinely in the Philippines. As The Economist noted in 2018, the partisanship of even senior office holders might be ambiguous: “To which political party does the president belong? In most countries, the question would have a straightforward answer, but not in the Philippines.” Most of the current legislators of the ruling party (Philippine Democratic Party–People’s Power, or PDP-Laban) belonged to other parties when they were elected, defecting after Rodrigo Duterte became the president in 2016. For the most part then, parties are not key actors in Philippine society. The dynamics whereby these parties partake in alliances with local violence specialists is nonetheless similar to that seen in Pakistan (and Nigeria). This creates a situation of “persistent but relatively low-level political violence” with much of the violence “directed either at intimidating or disposing of local political rivals, or at protecting particular economic enterprises” (Robinson Reference Robinson, Abraham, Newman and Weiss2010, 72). In the absence of ethnic linkages, the violence is not aimed at recruiting voters or polarizing the populace, and in the absence of strong parties, not often orchestrated directly by party organizations themselves. However, this pattern can turn to mass violence quickly where central government authorities seek to challenge the authority of local bosses and attempt to rein them in (Robinson Reference Robinson, Abraham, Newman and Weiss2010, 72).
In sum, multiparty democracy in the Philippines has been constrained by the presence of local, violent elites but individual parties and politicians have also capitalized on their presence for short-term personal benefit. Even in the absence of ethnic or other ascriptive identity cleavages, then, we see violence around election time and violence pervading local patron–client relations. As I suggest, ethnicity can be one organizing function through which polarizing politics plays out but is not necessary for the presence of party violence. In this case, the weak organizational structure of parties and the absence of a captive support base is critical to their alliances with violent bosses and other strongmen in landscapes of shared sovereignty.
8.4 The Shiv Sena and Direct Violence
The Shiv Sena is an organizationally strong, ethnic party in India that draws strong parallels to the Muhajir MQM in Karachi in its use of violence. The party was formed in 1966 in the megacity of Bombay (now Mumbai) in the state of Maharashtra by Bal Thackeray, a former journalist and cartoonist. Since his death in 2012, his son, Uddhav Thackeray, has taken over the mantle and also served as the Chief Minister of Maharashtra between 2019 and 2022.
The nativist Shiv Sena was formed as a “sons of the soil” movement that sought preferential economic policies for the local Marathi-speaking population, claiming that internal migrants, specifically South Indians, were taking an unfair share of jobs in the city. In the 1990s, as the party expanded to electoral arenas outside of the Mumbai–Thane beltFootnote 3 into the rest of Maharashtra, it shifted its discourse away from focusing only on the Marathi-speaking community to emphasizing a Hindu identity more broadly. Within Mumbai–Thane, however, the party’s regional appeal as a Marathi organization remains strong and coexists with Hindu-nationalist ideology. The Shiv Sena has been described as having a “ready capacity to incite widespread violence, extract rents, and shape public policy and legislative initiatives” (Katzenstein et al. Reference Katzenstein, Menhta and Thakkar1997, 371) and has relied on “aggression and everyday militancy right from its inception” (Bedi Reference Bedi2016, 6). Here, I briefly explain how the combination of the party’s organizational strength and the captive nature of its Marathi-speaking vote base sheds light on its capacity and incentives for violence in the particular political landscape of Mumbai.
The state of Maharashtra, with Bombay as its capital, was established in 1960, following demands for reorganization of Indian states along linguistic lines. A dramatic increase in educated Maharashtrian jobseekers in the early 1960s outpaced the number of available white-collar jobs, leading to fierce competition over limited clerical employment opportunities from South Indian migrants (Katzenstein Reference Katzenstein1979). This set the stage for the emergence of the Shiv Sena in 1966. In addition to changing demographics, Mumbai’s political landscape has other features in common with Karachi’s, even if state capacity is not lacking to the same degree. Hansen (Reference Hansen, Stepputat and Hansen2005) writes about Mumbai, for example, that “the right and the capacity to make decisions, to adjudicate, to govern, and even to kill and punish, historically has been distributed between a range of authorities and institutions in India” and that parties operate partly in “local and informal structures of authority and violence” (191).
The Shiv Sena – the term sena means army – was initially intended to be nonpolitical. It mobilized young men connected to youth clubs and gymnasiums in Marathi-dominated parts of the city and sought primarily to ostracize the “alien others believed to be dominating Bombay” while providing “security and sense of self-esteem … to young, frustrated males” (Hansen Reference Hansen2001, 48). In 1970, Joshi (Reference Joshi1970, 969) described the Shiv Sena as being “mid-way between a movement and a party” – a phrase that also applied to the MQM in its early years. While it began as an amorphous organization, upon entering politics soon after its inception, the Shiv Sena invested time in building a strong organization. It created a “three-tier structure” and selected and trained a “disciplined cadre of activists to man the various units” (Joshi Reference Joshi1970, 970). The middle tier, “the most crucial level in the structure,” was composed of “young militant activists” who were made chiefs of five parliamentary constituencies (Joshi Reference Joshi1970, 970). They were responsible for setting up branches in every municipal ward in their area and were involved in “criminal offenses and organized violence” (Joshi Reference Joshi1970, 974). The Shiv Sena relied heavily on shakhas, “small buildings with one or two fairly simple rooms containing chairs and often a picture or statue of Shivaji” (Hansen Reference Hansen2001, 53), which provided a daily meeting place for its activists. The shakhas were also seen as the place for citizens to go “for assistance and various forms of patronage” (Hansen Reference Hansen2001, 54).
Bal Thackeray’s primary position within the party hierarchy (he was the chief of Shiv Sena, or sena pramukh) was also critical to party organization, what Hansen calls the “dynamic of central charismatic authority” (2001, 60). He kept around him trusted and hand-picked individuals (the karya karani) who advised him on various policies. Decision-making was highly centralized and loyalty to the leadership was a key characteristic of party cadres. Rank-and-file members sought Thackeray’s attention and praise. Intraparty elections were rare, and no one was allowed to challenge his hold on power. Palshikar (Reference Palshikar2004) writes that Thackeray “never allowed any doubts to be raised as to his pre-eminent position in the party” (1500). Like the MQM, however, local party members were nonetheless imbued with significant authority. Above the shakha level, for example, was the zilla pramukh, or the district chief, who commanded activities in the district. Above him was the vibhag pramukh, or area commander, who acted as a liaison between party leadership and district chiefs. Cadres could steadily climb the party ranks, and many initial party cadres later served as senior leaders and members of legislative assembly (MLA). Shiv Sena activists are known as sainiks, a term that literally means soldier.
The organizational structure of the party in Mumbai–Thane remains robust today. A new class of leaders was introduced in the 1990s, known as the sampark pramukhs, or chief contact persons, responsible for each district (Palshikar Reference Palshikar2004, 1500); these individuals are, in turn, directly responsible to Uddhav Thackarey, who took over after his father’s death in 2012. Socialization of party cadres also continues to be emphasized. The party has long had a weekly cartoon known as Marmik through which “much of the political education of the Shiv Sena’s rank and file as well as its cadres takes place” (Joshi Reference Joshi1970, 971).
In its first decade of existence, the Shiv Sena represented a cross-section of Marathi speakers, drawn from across caste and income groups. Its explicit Marathi-first approach rested on the claim that Mumbai belonged to the Marathi people. Its appeal mixed addressing the needs of the Marathi community, including providing jobs for Marathi-speaking youth – “either through negotiations or intimidation” (Palshikar Reference Palshikar2004, 1499) – as well as “invoking the spirit of the Marathi identity” (Palshikar Reference Palshikar2004, 1501). The Sena, then, attracted both lower-middle class youth as well as a “section of Marathi intellectuals” who viewed the party as an extension of the movement that led to the formation of the Marathi-speaking state (Palshikar Reference Palshikar2004, 1502).
Following Hindu–Muslim riots in 1984 in Mumbai, Thane, and Bhiwandi, the Shiv Sena partly refashioned itself as a party also staunchly defending Hindu rights and fighting against Muslim aggression. This move coincided with an expansion into other electoral arenas in Maharashtra. Until his death, Bal Thackeray freely ridiculed Muslims, “calling them ‘rats,’ ‘poisonous snakes,’ landyas,Footnote 4 ‘traitors’ … and openly advocating that all Muslims go to Pakistan” (Hansen Reference Hansen2001, 86). This rhetoric was coupled with a sense that Thackeray spoke for the common man with a “simple moral code and sense of justice” (Hansen Reference Hansen2001, 59). Palshikar writes that the Shiv Sena’s appeal, then, hinges “on the adroit combination of nativism, regional identity and a communal construction of nationalism” (2004, 1501). Indeed, within Mumbai–Thane, the Shiv Sena maintains its exclusionary regional appeals and retains a largely captive Marathi-speaking support base. In 2017, for example, the Shiv Sena heightened its exclusionary Marathi rhetoric, denouncing its local competitor, the Hindu nationalist Bharatiya Janata Party, as being pro-Gujrati (Palshikar 2017).
From as early as 1967, the Shiv Sena has also engaged in routine and sometimes high-profile violence. This includes the murder of a Communist Party MLA in 1970, and attacks on opposition party offices, rallies, and South Indian restaurants (Katzenstein et al. Reference Katzenstein, Menhta and Thakkar1997, 374). A direct confrontation between the Sena and the state government in 1969 resulted in violent protests; hundreds of people were arrested and many were injured. According to Purandare (Reference Purandare1999, 121), this event showcased Thackeray and Shiv Sena’s ability to disrupt normal life in Mumbai. Hansen (Reference Hansen2001) describes this early use of violence by the Shiv Sena as “ritualized,” through which the identities of the sainiks was “affirmed and stabilized” (65) and part of a public performance which sought to highlight Maharashtrian cultural heritage. Engaging in violence therefore played a key role in further socializing Shiv Sena sainiks. Maheshwari (Reference Maheshwari2010, 9) quotes a sainik (and later MLA) as explaining the relationship between these activists and Thackeray in language strikingly similar to that used by MQM party members:
We are militant. We call ourselves sainiks so we obey our chief. Whatever he decides, we are ready to sacrifice our lives for the cause. He was the first one in India to do politics like this, without caring for consequences, and brought justice to the people of Mumbai. The strength of Bala Sahibji is such that he can paralyze anything in the city. The moment he puts his finger down and gives a go, everything can be started again.
The Shiv Sena’s use of violence is linked to an ideological position of “action-orientation,” whereby Sena workers are “exhorted to take action, be it social service, neighbourhood activity or strong-arm tactics and violence” in order to achieve the party’s goal and particularly in response to any perceived injustice against Marathi people (Palshikar Reference Palshikar2004, 1501). Supporters and members of the Sena see its use of violence as both the “inevitable effect of the apathy of the state government to the ‘just’ demands and expectations of the Marathi youth” as well as an opportunity for individuals to “settle private scores and make some private gains” (Palshikar Reference Palshikar2004, 1502). Its violence has also interacted with and amplified Mumbai’s reliance on dada-ism, a “metonym for a criminalized subculture” (Hansen Reference Hansen2001, 72), which refers to the role played by violent local strongmen in working-class communities as “protectors, brokers, and providers of services, jobs, and opportunities to men in the neighborhood.” Hansen writes that while the Shiv Sena did not create dada culture, “it brought this long-standing plebian discourse of masculine assertion into the public and political realms” (2001, 72). Many prominent dadas are members of the Shiv Sena, often serving in the role of zilla pramukh, and have been involved in street fighting and semilegal activities. Murder charges have also been filed against prominent Shiv Sena cadres. Bedi (Reference Bedi2016, 7) writes, “Over the course of my fieldwork, many low-level party workers, both men and women, told me they had been active participants in this violence in the mid-1990s.” Shiv Sena leaders and party members have also been arrested for suspected involvement in extortion.
Like Karachi, then, ethnic cleavages overlap with criminal networks and patron–client relations in Mumbai. The cadres’ role in solving neighborhood problems – whether directly or by going to local officials – and their involvement in local clientelism, constituency service, and general social welfare, as well as being seen as protectors of the community is key to their popularity. Shakhas, for example, provide an important community function, with some distributing schoolbooks and notebooks to needy children. The Shiv Sena has also frequently operated free ambulances and set up health camps or blood-donating drives (Hansen Reference Hansen2001, 105). During the Covid-19 pandemic, the party converted many of these shakhas into temporary clinics to treat patients (Mahamulkar Reference Mahamulkar2020). Its local-level presence allows for these activities, just as it permits it to operate and control violent specialists within its party apparatus.
As the Shiv Sena continues to seek to expand outside of Mumbai into other parts of Maharashtra state, we have seen signs – as this book would predict – that it has not adopted the same type of militant violence for which it has been known in Mumbai–Thane. Outside of this urban center, “the issue of sons-of-soil would not have clicked with the masses” (Palshikar Reference Palshikar2004, 1499), making it unlikely that the Shiv Sena could rely on a support base largely unfazed by – or indeed, supportive of – its use of violence. The Shiv Sena’s relative lack of organizational strength outside of Mumbai–Thane is another reason why such violence is not possible in the same way. As one scholar explained, “One reason for this difference [in violence strategies] between Mumbai-Thane and rest of the state is due to the fact that in Mumbai-Thane a more systematic cadre based network existed whereas in rest of the state, this has not been the case. There the party organization has been loose and chaotic – so orchestrated street violence was not possible, only emotive outbursts resulting in some clashes could be possible.”Footnote 5
The Shiv Sena’s role in Mumbai thus has many parallels with the MQM in Karachi. Its strong organizational structure permits it to engage in numerous forms of militancy, crime, and violence without having to rely on external actors, while its focus on a captive Marathi support bank ensures the absence of a cost from its primary constituency, with evidence of polarizing benefits as well. The party operates in an urban context where the potential gains from participation in an informal economy and through the production of violence are high. As the party seeks to expand its national appeal, we find that its violence strategies haven’t carried over to other electoral arenas in which it competes, both due to the lack of captive support base in these other locales and the relative weakness of its organizational capacity. As such, the case of the Shiv Sena also suggests commonalities with the ANP’s varied tactics in Karachi and KP province.
8.5 Conclusion
This book has presented a theory of party violence that originated and was tested in the Pakistani context. This chapter has sought to evaluate this book’s arguments by examining cases beyond that context. These comparisons have demonstrated compelling parallels with parties in Nigeria, the Philippines, and India, both in the strategies of violence that we see parties utilizing and the determinants of those strategies. In Nigeria, weakly organized parties outsource violence to armed criminal groups, often accompanied with exclusionary ethnic rhetoric, and also ally with locally powerful godfathers in areas of shared sovereignty. In the Philippines, if anything, the importance of local strongmen as electoral allies of parties is more central than it is in Pakistan and Nigeria. Finally, in India – which unlike Nigeria, the Philippines, and Pakistan, has never experienced military rule – the incentives for party violence still hold. The Shiv Sena engages directly in violence where politics are polarized along ethnic lines and where the party is organizationally most powerful, but as it has expanded to other arenas its use of violence has attenuated where voters are less captive and where the party’s infrastructure is newer and weaker. This analysis should, of course, be read as indicative but not definitive. While there are key differences in each case, they nonetheless demonstrate how a study of the MQM, ANP, PML-N, and PPP can help us understand why parties outside of Pakistan might use violence for electoral and nonelectoral gain, which parties are more likely to do so, and how voters are likely to react.
