• Skip to main content
  • Skip to primary sidebar
  • Skip to footer
  • Core Principles of Responsible AI
    • Accountability
    • Fairness
    • Privacy
    • Safety and Security
    • Sustainability
    • Transparency
  • Special Topics
    • AI in Industry
    • Ethical Implications
    • Human-Centered Design
    • Regulatory Landscape
    • Technical Methods
  • Living Dictionary
  • State of AI Ethics
  • AI Ethics Brief
  • 🇫🇷
Montreal AI Ethics Institute

Montreal AI Ethics Institute

Democratizing AI ethics literacy

Social media polarization reflects shifting political alliances in Pakistan

January 17, 2024

🔬 Research Summary by Alessandro Galeazzi and Fabiana Zollo.

Alessandro Galeazzi is a postdoctoral fellow at  Ca’ Foscari University of Venice, where he investigates infodemics and polarization on social media.

Fabiana Zollo, an Associate Professor of Computer Science at Ca’ Foscari University of Venice, where she leads the Lab of Data Science for Society and focuses on understanding the spreading and consumption of information, encompassing topics such as misinformation, social dynamics, and polarization within the digital realm.

[Original paper by Anees Baqir, Alessandro Galeazzi, Andrea Drocco, and Fabiana Zollo]


Overview: Increasing concern over ideological divisions in public discourse has grown in recent years. While political polarization has received extensive academic attention, regions such as South Asia and the Middle East have been understudied. To address this gap, this study explores the evolution of political polarization in online debates in Pakistan, a country marked by a complex interplay of military influence, regional conflicts, and religious extremism that recently experienced an overturn of the government, making it an interesting case for examination.


Introduction

Ideological divisions within society carry significant implications for social stability. Extreme polarization not only may foster the spreading of misinformation but also harbors the potential for dangerous consequences, as exemplified by events like the rise in vaccine hesitancy during the COVID-19 pandemic and the assault on Capitol Hill, which coincided with the most pronounced peak in political polarization in the United States since the Civil War.

While prior research often concentrates on Western nations, regions such as the Middle East and South Asia have received relatively little scholarly attention despite their growing global importance. To address this research gap, we analyze political polarization in Pakistan, using Twitter data spanning from 2018 to 2022. This period witnessed substantial political transformations, including Imran Khan’s coalition victory in 2018, the formation of an opposition alliance in 2020 that led to the ousting of his government in 2022, and Khan’s subsequent arrest in 2023.

Our analysis reveals a close alignment between Twitter debates in Pakistan and the unfolding of political events and evolving party alliances. Furthermore, it brings to light individual-level shifts, offering insights into the often-overlooked dynamics of Pakistani politics—a nation of considerable geopolitical significance. Our study contributes to a more comprehensive understanding of political polarization globally, examining its development within a socio-economic context that has frequently been disregarded in existing scholarly literature.

Key Insights

Online Polarization

Widening ideological divides pose a significant threat and represent a notable challenge to both social stability and the integrity of democratic processes such as elections, referendums, and the effective implementation of policies. While ideological polarization is a natural feature of pluralistic societies, its escalation to the extreme can jeopardize the collective well-being. This is evident in various events, from the COVID-19 vaccination campaign to heated climate change debates up to the shocking assault on Capitol Hill.  Hence, monitoring and understanding the causes and consequences of extreme ideological polarization is fundamental to preserving the stability and functionality of society and democratic institutions.

Studying Polarization in the Digital Age

Polarization is a longstanding phenomenon that has been subject to study through diverse methodologies, including analysis of voting records and surveys. Yet, online platforms have enabled researchers to exploit social media data for analyzing polarization in several contexts, like climate change, politics, and vaccines.  Most of these investigations have predominantly focused on Western countries, overlooking regions of political significance characterized by diverse socio-cultural dynamics. To address this gap, our research examines the evolution of polarization on Twitter within the Pakistani political landscape, offering a distinctive perspective on polarization within a non-Western, unique socio-cultural environment, enriching our understanding of this phenomenon in a broader global context.

The Pakistani Political Background

Since its establishment in 1947, Pakistan’s political history has been marked by persistent instability. The military has consistently influenced the nation’s politics, eroding trust in democratic institutions. Additionally, ethno-linguistic groups within the country seek recognition, leading to regional conflicts that challenge national unity. Religious extremism, exemplified by groups like the Taliban, further fuels polarization. In this environment, Pakistani media often display political biases instead of promoting impartial news coverage. In 2022, Pakistan’s Democracy Index ranking was notably lower than the threshold for full democracy, making it a significant case for understanding polarization in countries with different political systems. Furthermore, from 2018 to 2022, Pakistani politics experienced substantial changes, including the defeat of the ruling party and the arrest of the prime minister. The victory of the Pakistan Tehreek-e-Insaf (PTI, or Pakistan Movement for Justice in English) in the 2018 general election led to the formation of an opposition alliance, including the Pakistan Muslim League – Nawaz (PML-N) and the Pakistan People’s Party (PPP). Thanks in part to the defection of some PTI representatives, this coalition ultimately overturned the government in 2022, resulting in the arrest of Imran Khan in 2023.

Online Polarization Reflects Political Shifts

To investigate the unfolding of political events in Pakistan, we gathered Twitter (now X) timelines from members of parliament belonging to three major political parties: PTI, PPP, and PML-N, spanning the period from 2018 to 2022. Using this data, we reconstructed interaction networks based on retweets they received and measured the level of political polarization yearly. Our findings unveiled a robust connection between online polarization and the dynamics of Pakistani politics. This correspondence shed light on the coalescence of accounts and audiences affiliated with opposition parties, ultimately culminating in their convergence into one group in 2022.

Crucially, this convergence extended beyond party-level polarization to individual user accounts. The migration of defectors from the ruling party to the opposition was evident in online interactions, highlighting the sensitivity of techniques reliant on online data for estimating users’ ideological positions. Notably, as time progressed, the audiences of opposition parties grew more alike, while the online presence of PTI became increasingly isolated, with its content primarily consumed by its supporters. However, in 2022, there was a notable surge in new users exclusively supporting PTI, further exacerbating the polarization of the debate.

These results offer valuable insights into the interplay between online polarization and real-world political developments in Pakistan, shedding light on the state of ideological divisions in a non-Western nation and providing a vital perspective for a global understanding of polarization.

Between the lines

Recognizing the intricate relationship between online discourse and real-world occurrences is crucial, particularly in preventing potential challenges that could undermine societal cohesion, including the proliferation of extremism. Building on previous research, this paper underscores that social media data serves as a faithful reflection of offline events and can effectively serve as a tool for monitoring the trajectory of polarization, even in countries characterized by sociocultural contexts distinct from those predominantly explored in existing literature.

Polarization arises from multifaceted interactions among variables such as cultural divisions, economic disparities, and religious segregation, with its determinants being profoundly influenced by the specific environment. Since the efficacy of measures aimed at reducing radicalization and extremism may vary depending on the unique contextual factors at play, broadening the study of these phenomena to encompass diverse world regions holds paramount importance to gain a global perspective and develop tailored countermeasures to address these pressing challenges.

Want quick summaries of the latest research & reporting in AI ethics delivered to your inbox? Subscribe to the AI Ethics Brief. We publish bi-weekly.

Primary Sidebar

🔍 SEARCH

Spotlight

AI Policy Corner: Frontier AI Safety Commitments, AI Seoul Summit 2024

AI Policy Corner: The Colorado State Deepfakes Act

Special Edition: Honouring the Legacy of Abhishek Gupta (1992–2024)

AI Policy Corner: The Turkish Artificial Intelligence Law Proposal

From Funding Crisis to AI Misuse: Critical Digital Rights Challenges from RightsCon 2025

related posts

  • Rewiring What-to-Watch-Next Recommendations to Reduce Radicalization Pathways

    Rewiring What-to-Watch-Next Recommendations to Reduce Radicalization Pathways

  • Research summary: A Picture Paints a Thousand Lies? The Effects and Mechanisms of Multimodal Disinfo...

    Research summary: A Picture Paints a Thousand Lies? The Effects and Mechanisms of Multimodal Disinfo...

  • Is AI Greening Global Supply Chains?

    Is AI Greening Global Supply Chains?

  • Measuring Disparate Outcomes of Content Recommendation Algorithms with Distributional Inequality Met...

    Measuring Disparate Outcomes of Content Recommendation Algorithms with Distributional Inequality Met...

  • Mapping AI Arguments in Journalism and Communication Studies

    Mapping AI Arguments in Journalism and Communication Studies

  • On the Challenges of Deploying Privacy-Preserving Synthetic Data in the Enterprise

    On the Challenges of Deploying Privacy-Preserving Synthetic Data in the Enterprise

  • Research Summary: Towards Evaluating the Robustness of Neural Networks

    Research Summary: Towards Evaluating the Robustness of Neural Networks

  • Exploring Antitrust and Platform Power in Generative AI

    Exploring Antitrust and Platform Power in Generative AI

  • Does diversity really go well with Large Language Models?

    Does diversity really go well with Large Language Models?

  • Are we ready for a multispecies Westworld?

    Are we ready for a multispecies Westworld?

Partners

  •  
    U.S. Artificial Intelligence Safety Institute Consortium (AISIC) at NIST

  • Partnership on AI

  • The LF AI & Data Foundation

  • The AI Alliance

Footer

Categories


• Blog
• Research Summaries
• Columns
• Core Principles of Responsible AI
• Special Topics

Signature Content


• The State Of AI Ethics

• The Living Dictionary

• The AI Ethics Brief

Learn More


• About

• Open Access Policy

• Contributions Policy

• Editorial Stance on AI Tools

• Press

• Donate

• Contact

The AI Ethics Brief (bi-weekly newsletter)

About Us


Founded in 2018, the Montreal AI Ethics Institute (MAIEI) is an international non-profit organization equipping citizens concerned about artificial intelligence and its impact on society to take action.


Archive

  • © MONTREAL AI ETHICS INSTITUTE. All rights reserved 2024.
  • This work is licensed under a Creative Commons Attribution 4.0 International License.
  • Learn more about our open access policy here.
  • Creative Commons License

    Save hours of work and stay on top of Responsible AI research and reporting with our bi-weekly email newsletter.