The Impact of Hate Literature on Radicalization in Pakistan

How to cite

Mehmood, Sultan; Asad, Saher. 2021. "The Impact of Hate Literature on Radicalization in Pakistan." AEA RCT Registry. AEARCTR-0008251. https://doi.org/10.1257/rct.8251-1.0

Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.

Sultan Mehmood* , Saher Asad

* Primary investigator

Registry ID
AEARCTR-0008251
DOI
https://doi.org/10.1257/rct.8251-1.0
First registered
Last updated
Status
In development
Trial dates
2021-10-01 – 2022-08-10
Countries
Pakistan (South Asia)
Registry record
http://www.socialscienceregistry.org/trials/8251

Abstract

In many vulnerable democracies, radicalization poses a significant challenge. Radicalization as a precursor to political violence may be a threat to both democracy and development. Policymakers must decide from many policy options based on little to no evidence. Does limiting access to hate literature work best or is it the case that counter-narrative messaging and subsidizing competitors’ content more effective? We hope to inform this debate and contribute to the literature on determinants of political violence by conducting a pilot study where we explore the feasibility of several interventions that may reduce support for extremist ideologies. Our focus groups with vulnerable populations guide us to pilot 6 interventions. The randomized evaluation stemming from this pilot may provide new causal evidence on: i. how hate literature impacts support for extremist ideologies? ii. Does subsidizing counter-narrative messaging work, if so, what kind of counter-messaging is most likely to reduce support for extremist violence? iii. Does reducing access to hate literature and counter-narratives work as complements or substitutes?

Intervention

In this pilot, we roll out treatment across 35 villages with 5 villages per treatment (with 18 individuals per village). Our individual level sample for 35 villages will collect baseline and end line information for a total of 630 individuals in the 35 villages. The data from pilot will then also be used for conduct power calculations for the full experiment and assess feasibility of some of the interventions. For instance, the pilot will allow us to assess the response of readers of hate literature to our supply shock of removing these magazines from main display shelves or relegating them to “middle invisible shelf” (according to a bookshop owner’s suggestion). The pilot will also help us further in understanding which interventions are feasible as well as most interesting and have potential. The assignment to treatment will be done at the village level but the outcomes will be measured at the individual level. To summarize, we hope to the following pilot 6 treatments and 1 control across the 35 villages and implement the following interventions:

T1: Invisible Hate Treatment – Paying bookshop owners not to display extremist publication.
T2: Counter-Narrative Treatment – Introducing popular counter-narrative literature without changing supply:
T2a: Secularization Counter Messaging – we subsidize secular positive science publication while keeping the supply of Zarb e Momin (hate literature) unchanged.
T2b: Religious Counter Messaging - a religious non extremist treatment while keeping the supply of Zarb e Momin unchanged.
T3: Education Treatment - Educating people on responsible media use and impacts of hate speech.
T4: Negative Access to Hate Shock X Counter-Narrative– T1 and introducing counter-narrative
T5: Negative Access to Hate Shock X Education – T1 and educating people on responsible use of media and impacts of hate speech
C: Control Group – No intervention

Experimental design

We will design and pilot baseline and endline (phone & in-person, depending on the COVID-19 situation) surveys with people in sample villages. Our primary population of interest will be a random sample of individuals in age range 18-35 years old in treatment and control villages. We also plan to stratify on gender as men and women have different channels of information so a comparison of men and women will provide useful insights. We will pilot the following treatments which we randomize at the village level:
T1: Invisible Hate Treatment – Paying bookshop owners not to display extremist publication.
T2: Counter-Narrative Treatment – Introducing popular counter-narrative literature without changing supply:
T2a: Secularization Counter Messaging – we subsidize secular positive science publication while keeping the supply of Zarb e Momin (hate literature) unchanged.
T2b: Religious Counter Messaging - a religious non extremist treatment while keeping the supply of Zarb e Momin unchanged.
T3: Education Treatment - Educating people on responsible media use and impacts of hate speech.
T4: Negative Access to Hate Shock X Counter-Narrative– T1 and introducing counter-narrative
T5: Negative Access to Hate Shock X Education – T1 and educating people on responsible use of media and impacts of hate speech
C: Control Group – No intervention

Design details

Randomization method
Village level randomization done by a computer (in Stata).
Randomization unit
Village
Clusters
Pilot will have 35 villages with 18 individuals per village.
Observations
18 x 35 = 630
Arms
7 arms including control so 35/5 = 5 villages per treatment arm for this pilot
Primary outcomes
The key list of outcomes we plan to study include conducting surveys in the pilot: terrorism incidents in the district, support for extremist ideologies; support for terrorist organizations; willingness to pay for accessing radical literature; number of terrorist cases registered in the area, perspectives on other social issues such as women’s empowerment (domestic violence, women activism for their rights such as right to an inheritance), support for political violence, financial support for terrorist organizations, willingness to pay for accessing radical literature perspectives on women’s empowerment, degree of slant in gender norms and other development outcomes such as self-reported income and asset ownership. We will conduct baseline and endline surveys by phone whenever possible. The baseline survey we will be conducted before starting the interventions and endline survey will be conducted 6 months after launch of interventions.

Cited by

No citation data available for this registration yet.

Keywords