Political Parties and Violence in Karachi, Pakistan

• Author: Niloufer Siddiqui
• DOI: http://doi.org/10.1177/00104140221115174
• Published date: 01 April 2023
• Date: 01 April 2023
• Subject Matter: Articles

Article

Comparative Political Studies

2023, Vol. 56(5) 726–756

© The Author(s) 2022

Article reuse guidelines:

sagepub.com/journals-permissions

DOI: 10.1177/00104140221115174

journals.sagepub.com/home/cps

Political Parties and

Violence in Karachi,

Pakistan

Niloufer Siddiqui

1



Abstract

Political parties in many ethnically polarized contexts around the world

engage in electoral violence. Why do some parties engage in violence directly

while others do so through intermediaries? Existing literature has produced

contradictory results about when democracy turns violent and paid insuff‌i-

cient attention to the form that party violence takes. Using qualitative and

survey data of both party elites and voters, this paper focuses on two parties in

the violent megacity of Karachi, Pakistan: the Muttahida Qaumi Movement—

whose cadres directly orchestrated violence—and the Pakistan People’s

Party—which outsourced violence to distinct specialists. It argues that where

state capacity is contested by multiple competing sovereigns, the economic

and coercive benef‌its to be gained from violence are high. However, even

where parties share similar incentives for violence within a particular electoral

arena, whether a party engages in violence directly or not depends on its

organizational structure and strength.

Keywords

conf‌lict processes, elections, public opinion, and voting behavior, political

parties, qualitative methods

1

University at Albany, State University of New York, Albany, NY, USA

Corresponding Author:

Niloufer Siddiqui, University at Albany, State University of New York, 135 Western Ave, Albany,

NY 12203, USA.

Email: nasiddiqui@albany.edu

In 2010, Raza Haider, a legislator belonging to the Muhajir-centric

1

Muttahida

Qaumi Movement (MQM) political party, was shot and killed during a funeral

in Karachi, Pakistan. For hours following his killing, armed men roamed the

city seeking revenge, many draped in the MQM party f‌lag. By that evening, 46

were dead and more than 100 injured (Imtiaz & Walsh, 2010). While this

incident was particularly noteworthy—Haider was, after all, a member of the

Sindh provincial assembly—ethnic violence in this megacity of more than 15

million had long ceased to surprise its residents. The MQM’s party cadres had

a reputation for their militant approach to politics, with widely believed al-

legations that they had maintained a torture cell for their opponents and party

dissidents.

Five years later, a confession implicated another political party for a role in

Karachi’s violence. Uzair Baloch, notorious gangster and head of the People’s

Aman Committee (PAC)—a Karachi-based ethnic militia—admitted to

having engaged in illegal acts and criminal activities on behalf of leaders of

the Pakistan’s People’sParty (PPP), with some news reports even alleging that

he had carried out assassinations at their behest. The PPP had relied on the

PAC to secure them valuable votes in Karachi, address constituent demands,

and, so it seemed, orchestrate violence at their orders. Outside of Karachi, the

PPP, while roundly criticized for corruption, was not known to be a violent

party.

While both events reveal a link between political parties and violence, they

offer quite different relationships between the party and the violent act. The

MQM developed a reputation for direct violence, while the PPP opted to

outsource violence to the PAC. This raises the question: Why do some parties

engage in violence directly through their own party cadres while others

outsource the task to distinct armed actors? The answer is not obvious from

f‌irst principles. Being in control of your own violence-wielding apparatus

provides distinct advantages over relying on external specialists. In particular,

it helps minimize principal-agent problems such that loyal party cadres act in

accordance with their leadership’s commands regardless of whether it is in

their personal best interest; in contrast, violence specialists “often defy,

manipulate, and bargain with”(Staniland, 2015b, p. 694) the politicians on

which they otherwise rely. On the other hand, unlike in the process of out-

sourcing where parties are able to sever links with violence specialists when

needed, parties carrying out violence directly are less able to distance

themselves from the violence and hence, less able to benef‌it from plausible

deniability.

Answering this question is not merely of interest to scholars of Pakistan.

Ethnic violence is a feature of many democracies around the world, new and

well-established alike. Growing evidence from Africa and Asia to Latin

America increasingly points towards political parties engaging in coercion

and outright violence as especially notable forms of electoral irregularities

Siddiqui 727