Election forecasts, based on public opinion polls or statistical structural models, regularly appear before national elections in established democracies around the world. However, in less established democratic systems, such as those in Latin America, scientific election forecasting by opinion polls is irregular and by statistical models is almost non-existent. Here we attempt to ameliorate this situation by exploring the leading case of Argentina, where democratic elections have prevailed for the last thirty-eight years. We demonstrate the strengths—and the weaknesses—of the two approaches, finally giving the nod to structural models based political and economic fundamentals. Investigating the presidential and legislative elections there, 1983 to 2019, our political economy model performs rather better than the more popular vote intention method from polling.
Argentinian Elections
