The paper describes a general approach to the modelling of nonlinear a
nd nonstationary economic systems from time-series data. This method e
xploits recursive state space filtering and fixed interval smoothing a
lgorithms to decompose the time-series into long term trend and short
term 'small perturbational' components, each of which are then modelle
d by linear stochastic models which may be characterised by time varia
ble parameters. The approach is illustrated by an example which explor
es the relationship between the variations in quarterly GNP and Unempl
oyment in the USA over the period 1948 to 1988 and questions previous
claims about the changes in the slope of the long term trend in log(e)
(GNP) over this same period. The paper also points out that the recurs
ive approach to estimation facilitates the use of these methods in the
development of adaptive forecasting and control systems.