ABSTRACT
This chapter has two main concerns. It addresses generally the problem of short-term forecasting in international affairs. First, the difficulties of providing systematic and useful forecasts about environments that are characterized by uncertainty and rapid change are noted. Consideration then turns to the qualities desired in short-term international forecasts and the necessary capabilities of a methodology used to produce such forecasts on a real-time basis. It is argued that the specifications for such forecasting require the application of certain types of stochastic models that incorporate Bayesian methods to allow integration of information from sample data with subjective estimates obtained from expert analysts.
