A time varying parameter approach to the Chinese aggregate consumption function

China experienced a number of economic, political and social upheavals in the pre-reform period, together with a gradual transformation from a centrally-planned to a market oriented economy in the post-reform period. Given this background of extensive change, a time varying parameter (TVP) consumption model for non-durables is developed in order to determine the resulting changes in consumer behaviour caused by both observable and unobservable factors. The parameters of interest are the short and long run marginal propensities to consume (MPC) and the long run average propensity to consume (APC). The model is based on Friedman's permanent income hypothesis (PIH) and estimated using the Kalman filter algorithm. The empirical results suggest that the TVP model is a good representation of the changes in Chinese consumer behaviour over time. In terms of forecasting, the TVP model generally outperforms a number of alternative models.