Erikson and Welzien: Of Time and Presidential Election Forecasts
Essay title: Erikson and Welzien: Of Time and Presidential Election Forecasts
Of Time and Presidential Election Forecasts clearly indicates by the authors will attempt to explain the relation between time and election forecasts. The introduction tells you more specifically, that we are interested in just how far in advance electoral expectations come into focus. The article is very clear an appropriately written for its target audience.
This article examines a model of predicting an election that has been successful at predicting past presidential elections. The two variables are cumulative per capita income growth and presidential approval. These “fundamental” variables can predict the results well when measured shortly in advance of the election, when the outcome is already becoming clear in the polls. However, as one steps further back from the election, the predictive power is depleted. Income growth and approval readings taken early in the election year can barely predict readings of the same variables just prior to the election.
The findings reviewed show that approval is a stronger indicator, but also that as the electoral cycle evolves, the economy influences the Presidents standing which, in turn, reflects the verdict in the ensuing election in advance.
The findings show that presidential elections are basically unpredictable until early in the election year itself, but then preferences harden rather quickly. The public continues to update its preferences as the election approaches.
Forewarned before Forecast: Presidential Election Forecasting Models and the 1992 Election indicates that we will assessing election forecasting models, but does not give us an indication of how. The realibility of the forecasts is tested using models by James Campbell, Alan Abramowitz, Michael Lewsi-Beck, Rice and Ray Fair. Only Campbells model came up with an acceptable measure of the confidence interval indicating that the models were based on too few observations to be precise. An assessment of model composition showed that the Beck, Rice and Abramowitz models were similar in principle