In this paper, we estimate a consumption function based on a new set of data for household wealth. The basis is a standard error correction model where consumption is driven by income and wealth developments in the long run. The model which is estimated on quarterly data for the period 1973-2005 is characterized by a long-run relationship which is significant and stable over time. Moreover, the model satisfies a homogeneity requirement, implying that consumption is proportional to income and wealth in the long-run. The model can, basically, be interpreted as a re-estimated version of the existing consumption function in Mona, which is based on income and wealth data for the aggregate private sector. However, the latter model has difficulties in explaining consumption developments in the years after 1998 where the relationship between consumption and wealth ratios for the private sector has turned out not longer to be valid. The use of data for households is theoretically more adequate when modelling private consumption and, at the same time, solves the problem of model instability over the most recent years.
consumption function household wealth model stability