The purpose of this article is to explore electoral forecasting and public opinion tracking in Latin America. We review different approaches used to estimate the true value of public opinion, and assess the range of their application. We focus on election night forecasting and campaign variation tracking in Latin America. We propose a two-stage model based on poll aggregation and Bayesian inference. We present data from two presidential elections in Chile. We test the model and show that it provides the most accurate election night forecasting point estimate and the most comprehensive campaign variation tracking method. Finally, we discuss the advantages and limitations of our
model, and suggest a route for future research.
Bunker, K., & Bauchowitz, S. (2016). Electoral Forecasting and Public Opinion Tracking in Latin America: An Application to Chile. Política. Revista De Ciencia Política, 54(2), pp. 207–233. https://doi.org/10.5354/0719-5338.2016.44781
Twitter