A Re-evaluation of Housing Wealth Effect in Korea

This paper attempts to re-evaluate the size of housing wealth effect in Korea. Our focus is on the size of ‘genuine’ housing wealth effect, i.e., the response of consumption spending by home-owners to the changes in housing wealth. Two issues show up while we estimate the ‘genuine’ wealth effects using aggregate time series data: the issues around home ownership and proper measure of consumption. We first argue that it is more appropriate to use non-housing consumption, because housing consumption is in large part not of the choice of home owners but the imputed rents they do not actually choose to pay. We then proceed to address the issue of home ownership, by examining how much to revise the estimates of housing wealth effect obtained from aggregate non-housing consumption data. We construct two structural models and estimate the share of home-owners' consumption in those models' context. It is found that, if properly revised in light of the estimated consumption shares of home-owners, the magnitude of resulting housing wealth effects are larger than what simple time series regressions imply