Consumption and income in Finland 1960-1983 : A Multiple time series analysis

Abstract

The aim of this paper is to find out whether there are gains with respect to forecasting from modelling consumption and income as a bivariate time series process rather than as univariate processes. It is found that consumption and income can be modelled best jointly through an ARMA2(2,l)model. Income is found to be Granger causally prior to consumption.