First-time Voters, Lasting Democracy
How to cite
Sukhtankar, Sandip; Mobarak, Ahmed; Kosec, Katrina; Mahzab, Moogdho. 2026. "First-time Voters, Lasting Democracy." AEA RCT Registry. AEARCTR-0017876. https://doi.org/10.1257/rct.17876-1.1
Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.
* Primary investigator
Abstract
This study investigates how decentralized political movements built around street protests transform into formal political forces, specifically within the context of Bangladesh’s democratic transition following the July–August 2024 student-led uprising. Despite the critical role played by women and youth in toppling the previous authoritarian regime, these groups have remained largely marginalized in the subsequent interim government and political leadership. This project utilizes a randomized controlled trial (RCT) in rural Bangladesh to test whether video narratives highlighting the contributions of these underrepresented groups can influence political engagement in the February 2026 national elections.
The intervention involves community screenings combined with door-to-door and digital information campaigns deployed at the polling-station level. Clusters are randomly assigned to one of four arms: (1) a video narrative highlighting the role of male student protesters in the 2024 uprising; (2) a video narrative highlighting the role of female student protesters and organizers; (3) a video narrative focusing on the historical 1971 independence struggle; and (4) a pure control group. By leveraging administrative election data and exit polls, the study estimates the causal effect of these narrative framings on electoral turnout and voting behavior. The findings aim to provide insight into how narrative framing shapes democratic inclusion in fragile and transitional political settings.
Intervention
Polling stations are randomly assigned to one of four groups
Treatment 1: Male-Centered Protest Narrative (Contemporary)
Intervention: A 5-minute video highlighting the July–August 2024 student-led uprising, featuring clips and narratives centered on male student leaders and protesters.
Treatment 2: Female-Centered Protest Narrative (Contemporary)
Intervention: A 5-minute video mirroring the structure of Treatment 1 but substituting imagery and narratives to center female student leaders and organizers.
Treatment 3: Historical Narrative
Intervention: A 5-minute video focusing 1971 independence struggle, secular narrative, and religious harmony.
Control Group
Intervention: No video intervention.
Experimental design
The study employs a clustered randomized controlled trial (RCT) to evaluate the impact of narrative video interventions on voting behavior in rural Bangladesh. The experiment is timed to precede the February 2026 national election. The design tests whether highlighting the specific contributions of marginalized groups (women and youth) in the 2024 uprising affects their subsequent political participation and the electorate's perception of their legitimacy.
Design Overview
Unit of Randomization: The randomization unit is the polling station catchment area, which typically corresponds to a distinct village boundary. This minimizes spillover effects between treatment and control groups.
Stratification: Randomization is stratified first by political constituency to ensure geographic balance. Within constituencies, polling stations are stratified based on female voter share, as we anticipate that women's share of the population and of registered voters in particular may moderate the impacts of treatments on turnout.
Treatment Arms: Polling stations are randomly assigned to one of four groups:
Treatment 1: Male-Centered Protest Narrative (Contemporary)
Treatment 2: Female-Centered Protest Narrative (Contemporary)
Treatment 3: Historical Narrative
Control Group
Intervention Delivery
The intervention is delivered via a hybrid model to ensure broad coverage within the treated clusters:
Door-to-Door: Field staff visit households to show videos directly to residents.
Community Screenings: Videos are shown in public gathering spots
Digital Dissemination: Links to the videos are shared via WhatsApp to residents within the catchment area.
Data Collection & Outcomes Data will be collected from two primary sources:
Administrative Data (Election Commission): Official turnout rates and vote shares at the polling-station level.
Day-of-election data: Conducted on election day (12 February 2026) to capture turnout by demographic categories, including gender, perceived religiosity, and age group.
Primary Outcomes: Voter turnout, vote choice.
Design details
Cited by
No citation data available for this registration yet.