Why polls fail to predict elections

Abstract

In the past decade we have witnessed the failure of traditional polls in predicting presidential election outcomes across the world. To understand the reasons behind these failures we analyze the raw data of a trusted pollster which failed to predict, along with the rest of the pollsters, the surprising 2019 presidential election in Argentina. Analysis of the raw and re-weighted data from longitudinal surveys performed before and after the elections reveals clear biases related to mis-representation of the population and, most importantly, to social-desirability biases, i.e., the tendency of respondents to hide their intention to vote for controversial candidates. We propose an opinion tracking method based on machine learning models and big-data analytics from social networks that overcomes the limits of traditional polls. This method includes three prediction models based on the loyalty classes of users to candidates, homophily measures and re-weighting scenarios. The model achieves accurate results in the 2019 Argentina elections predicting the overwhelming victory of the candidate Alberto Fernández over the incumbent president Mauricio Macri, while none of the traditional pollsters was able to predict the large gap between them. Beyond predicting political elections, the framework we propose is more general and can be used to discover trends in society, for instance, what people think about economics, education or climate change.