his study empirically investigates the factors affecting local real estate taxation in 19 municipalities in Taiwan. The factors are determined using panel data over the period from 2001 to 2018. In several cases, the estimates obtained through quantile regression are notably different from those obtained through ordinary least squares (OLS) regression. The quantile regression of per capita gross domestic product (GDP) reveals a significant and positive influence of GDP on the lower tail of deed taxation and land value increment taxation. Furthermore, the quantile regression estimates have a positive influence on the middle and higher tails of house taxes and on the lower and higher tails of land value taxes.
The quantile regression estimates are significant in the upper tail of the real estate holding tax levied on the transfer of building ownership. The results of this study confirm that quantile regression estimates of real estate taxes are notably different from those obtained through OLS regression. Therefore, some policies related to local real estate taxation should be reconsidered, especially those pertaining to the adjustment ratios of announced land value, current land value, and standard unit housing prices.