2025/07/01 by Francesca R. Jensenius, Pradeep Chhibber, Sanjeer Alam +2 · 1 voice
Social Sciences · #Electoral Systems and Political Participation #Southeast Asian Sociopolitical Studies #Media Influence and Politics
paper · pdf · doi:10.1038/s41597-025-05418-6
openalex publication_date 2025/07/01 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/23
This paper describes the creation of a dataset covering polling-station-level election returns from the 2009, 2014, and 2019 Indian general elections. Indian elections are the largest in the world, and parliamentary constituencies are enormous. Within-constituency analysis is therefore crucial to address important political questions. Polling stations are the lowest level at which election returns are released. We have collected, cleaned, and standardized data on votes cast for each candidate across about 500,000 polling stations in 11 states for 2009 and 2014 and about 162,000 polling stations from one state for 2019. Polling-station numbers change between elections, while census units are more stable and can be traced over time. Using census indicators, we manually link polling stations to villages and towns in their catchment area, making it possible to merge the election data with census data and to observe local-level voting patterns over time. Our manual approach enabled us to link about 95% of the polling stations to corresponding census units and ensured a high degree of data reliability.