A Study on the Impact of Living Floor on the Price of an Apartment using a Spatio-temporal Model

This study analyzes the impact of living floor on the price of an apartment using 3,249 price data sample of apartments traded in Sejong city in 2017. In addition to a traditional regression analysis, we employ a spatial lag model, and a spatio-temporal model to consider the prices of neighboring apartments. Among these, spatio-temporal model turned out to be the fittest since its log likelihood was the biggest with having the smallest AIC and SC value. In regard to living floor which is our main concern, we made a dummy variable per 3- floor- unit with having 1-3 floor unit as our reference variable, and the highest dummy being 28-30 floor unit dummy. We controlled such variables as size, total floor, age, squared age, the number of households, parking lot capacity, stair type apartment dummy, distance from government complex, distance from Daejon city hall, BTR dummy, brand name dummy, and time fixed effect dummy which can also affect the price. As results, we found that except for the number of households variable and time fixed effect variables, most variables were significant with 1% level. Especially, living floor dummy of our interest showed that except for 19-21 floor dummy, the higher the floor, the higher the price. This result is different from literature where living floor raises the price up to a certain point of floor, but after the point the price rather declines. This may be because the residents in Sejong city are comparatively young, on average, and they may favor high living floors which can provide good environmental benefits. Besides, the coefficient of relative living floor (living floor/total floor) was bigger than that of living floor, which shows that apartment purchasers value relative height of living floor more than just the height of living floor.