Determinants of house prices in Seoul: A quantile regression approach

This paper estimates house price determinants in the Korean housing market, focusing on Seoul and employing the method of a quantile regression of a hedonic price model. With this quantile regression approach, how implicit prices of housing characteristics vary in each quantile of house prices is investigated. The data sample for this paper comprises court auction data collected from district courts of Seoul between 2006 and 2012. The hedonic variables employed in this research include building age, size, floor height, and floor level, proximity to metro station and high school and scenic view. As dependent variables, house prices in three sub-regions of Seoul are considered. Those prices between before and after the 2008 financial crisis are compared to determine the impact of the crisis. The empirical analysis finds that school proximity has the largest effect on the prices among dummy variables and that the level of the effect is larger in lower quantiles (lower-priced houses). By contrast, scenic view has a larger effect in higher quantiles (higher-priced houses), while its effect is statistically insignificant in lower quantiles. The effects of both school proximity and scenic view are also found to decrease in magnitude after the crisis.