Introduction

Forecasting is clearly crucial to such a process. Planning is a process of analysis and action which is necessarily about the future. It involves intervention to manipulate procedures or activities in order to achieve goals. The formulation of a typical planning model can be considered in five stages: 'selection of the variables to be included; choice of appropriate level of aggregation and categorization; decisions about the treatment of time; specification; and calibration.' Almost all practical planning models are static or comparative static equilibrium models that are they deal with discrete observations in time and operate as though these represent equilibrium states of the subsystem, which can then be easily extrapolated to provide forecasts. By contrast, abstract models represent the relevant characteristics of the original in a language, which might be expressed verbally using words and syntax or mathematically using symbols and equations.