Do house prices affect consumption and why? A replication and comparison exercise∗

Campbell and Cocco (2007) and Attanasio, Blow, Hamilton, and Leicester (2005a) (CC and ABHL hereafter) look at the effect of house prices on consumption in the UK and arrive at two opposite conclusions. CC compares groups of consumers classified by age-homeownership status and find that old owners are the most benefited by a house price increase and young renters the least. They interpret these findings as supportive of the wealth hypothesis. ABHL follow a different route and compare consumers by age status. These authors find that house prices have the same impact on consumption across age groups, which according to the authors suffices to discard the wealth channel hypothesis in favor of the common factor hypothesis. This paper performs a comparison exercise between the two papers to see what is the driving factor that causes both papers to reach these opposite conclusions. Special attention is given to the dimension (age or home ownership) that is used by each paper to test these two hypothesi. We first show that, once ABHL’s specification has been fully modified to replicate that of CC, the wealth hypothesis continues to be challenged across the age dimension: The consumption of the young and old do not react significantly different to a house price increase. Results from the comparison across ownership groups (owners and renters), and across age-ownership status groups are however not robust to the different specifications used. To the extent that there are differential effects of house prices on consumption depending on age-home ownership status, the fact that the conclusions depends on the specification used calls for a more rigorous theoretical discussion on the right empirical specification that is needed to test the wealth vs. the common factor hypothesis