Forecasting Spanish elections

Abstract

The behavior of the individual Spanish voter has come to be rather wellunderstood, thanks to a growing research literature. However, no models have appeared to explain, or to forecast, national election outcomes. The presence of this research gap contrasts sharply with the extensive election forecasting work done on other leading Western democracies. Here, we fill this gap. The model, developed from core political economy theory, is parsimonious but statistically robust. Further, it promises considerable prediction accuracy of legislative and European election outcomes, six months before the contest actually occurs. After presenting the model, and carrying out extensive regression diagnostics, we offer an ex ante forecast of the November 2011 legislative election, and then discuss the model and the forecast in light of what turned out to be the actual election result.