Wikipedia traffic data and electoral prediction: towards theoretically informed models

TL;DR

This paper develops a theoretical model to understand how Wikipedia page view data can predict electoral outcomes, especially changes in turnout and vote share, by analyzing information-seeking patterns related to different party types.

Contribution

It introduces a theoretical framework linking online information seeking to electoral results and tests it on European election data, improving understanding of social media data's predictive mechanisms.

Findings

Wikipedia data predicts changes in turnout and vote share.

Different party types generate distinct information-seeking patterns.

Theoretical model explains online activity's relation to electoral outcomes.

Abstract

This aim of this article is to explore the potential use of Wikipedia page view data for predicting electoral results. Responding to previous critiques of work using socially generated data to predict elections, which have argued that these predictions take place without any understanding of the mechanism which enables them, we first develop a theoretical model which highlights why people might seek information online at election time, and how this activity might relate to overall electoral outcomes, focussing especially on how different types of parties such as new and established parties might generate different information seeking patterns. We test this model on a novel dataset drawn from a variety of countries in the 2009 and 2014 European Parliament elections. We show that while Wikipedia offers little insight into absolute vote outcomes, it offers a good information about changes in both overall turnout at elections and in vote share for particular parties. These results are used to enhance existing theories about the drivers of aggregate patterns in online information seeking.