Polarization and acculturation in US Election 2016 outcomes – Can twitter analytics predict changes in voting preferences

Purva Grover, Arpan Kumar Kar, Yogesh K. Dwivedi, Marijn Janssen

Research output: Contribution to journalArticleScientificpeer-review

81 Citations (Scopus)
12 Downloads (Pure)


Elections are among the most critical events in a national calendar. During elections, candidates increasingly use social media platforms to engage voters. Using the 2016 US presidential election as a case study, we looked at the use of Twitter by political campaigns and examined how the drivers of voter behaviour were reflected in Twitter. Social media analytics have been used to derive insights related to theoretical frameworks within political science. Using social media analytics, we investigated whether the nature of social media discussions have an impact on voting behaviour during an election, through acculturation of ideologies and polarization of voter preferences. Our findings indicate that discussions on Twitter could have polarized users significantly. Reasons behind such polarization were explored using Newman and Sheth's model of voter's choice behaviour. Geographical analysis of tweets, users, and campaigns suggests acculturation of ideologies among voting groups. Finally, network analysis among voters indicates that polarization may have occurred due to differences between the respective online campaigns. This study thus provides important and highly relevant insights into voter behaviour for the future management and governance of successful political campaigns.

Original languageEnglish
Pages (from-to)1-48
JournalTechnological Forecasting and Social Change
Publication statusPublished - 2018


  • Acculturation in social media
  • eParticipation
  • Hashtag community
  • Polarization in elections
  • Public policy
  • Social media
  • Twitter analytics


Dive into the research topics of 'Polarization and acculturation in US Election 2016 outcomes – Can twitter analytics predict changes in voting preferences'. Together they form a unique fingerprint.

Cite this