Abstract
The use of socio-technical data to predict elections is a growing research area. We argue that election prediction research suffers from under-specified theoretical models that do not properly distinguish between ‘poll-like’ and ‘prediction market-like’ mechanisms understand findings. More specifically, we argue that, in systems with strong norms and reputational feedback mechanisms, individuals have market-like incentives to bias content creation toward candidates they expect will win. We provide evidence for the merits of this approach using the creation of Wikipedia pages for candidates in the 2010 US and UK national legislative elections. We find that Wikipedia editors are more likely to create Wikipedia pages for challengers who have a better chance of defeating their incumbent opponent and that the timing of these page creations coincides with periods when collective expectations for the candidate's success are relatively high.
Acknowledgement
The authors would like to thank David Rohde for his generosity in supplying data on candidate quality.
Disclosure statement
No potential conflict of interest was reported by the authors. The views and conclusions contained herein are those of the authors and should not be interpreted as necessarily representing the official policies or endorsements, either expressed or implied, of IARPA, DoI/NBE, or the US Government.
Notes on contributors
Drew Margolin is an assistant professor in the Department of Communication at Cornell University. His research focuses on the mechanisms that shape the production of individual and institutional discourse on the Internet and social media.
Sasha Goodman is a post-doctoral research fellow in the Political Science Department at Northeastern University and an affiliate of the Harvard Institute for Quantitative Social Science. His research focuses on organizational theory and political organizations.
Brian Keegan is a post-doctoral research fellow at in the Political Science Department at Northeastern University and an affiliate of the Harvard Institute of Quantitative Social Science. He researches social networks and online collaboration using computational social science.
Yu-Ru Lin is an assistant professor in the School of Information Sciences at the University of Pittsburgh. Her work focuses on social and political networks as well as computational and visualization methods for understanding network data.
David Lazer is a professor in the Political Science Department and the College of Computing and Information Science at Northeastern University and a Visiting Scholar at the Harvard Kennedy School. He works in the intersection of computational social science, political science, and social networks.