ABSTRACT
This chapter describes a model which is simple enough to provide some precise quantitative results and which can be used as the starting point for a step by step expansion to more complex models. This leads to a three equation model with non-stochastic explanatory variables and well behaved error terms. The problem of coefficient constraints will arise when the constraint item (CI) equation does not contain every explanatory variable in the model. The determination of the some factors should provide insights as to when it might prove beneficial to choose a CI equation with excluded explanatory variables. The CI forecasting procedure appropriate to this model is as follows: The coefficients of equations not chosen as the CI are estimated by ordinary least squares. The forecast of the CI is then determined as the value which, together with the forecasts of the other dependent variables, satisfies the adding up constraint.
