Index Methods for Forecasting: An Application to the American Presidential Elections

Abstract

Scott Armstrong and Alfred Cuzan describe Allan Lichtman's Keys Model as an example of an index method of forecasting, which assigns ratings of favorable, unfavorable, or indeterminate to influencing variables. They describe how index methods have been applied in other decision-making contexts, and they discuss when such methods might be useful analytical tools for business forecasters. In the context of presidential election forecasting, they compare the keys model to several regression models and find that it stacks up quite well against these more sophisticated alternatives.

JEL Classification: COO, C2