Diffusion of House Price Dynamics in Tehran: Using Spatial Autoregressive and Vector Error-Correction Models

Few studies have recently supported the preposition that housing price changes in one location can be affected not only by its history but also by housing price changes in other locations (Ripple effect). There are many reasons why housing price changes in some areas may lead housing price movements in the other areas. The possible causes include structural differences and economic interdependence between areas, migration as well as informational factors. This paper examines spatial effects of house price dynamics within Tehran. The first hypothese is that house prices have started rising first in the north and then spread out over the rest of the city. The second hypothesis is that there is a strong price relationship between contiguous districts. Spatial autoregressive (SAR) and vector error-correction (VEC) models are emplyed to estimate the empirical models using quarterly panel data of Tehran housing markets over the period 1991 to 2006. The results confirms that there is a positive spatial interrelationship among districts. Moreover, the findings indicate the evidence of leading role of the district 1 in the northern half of the city while district 2 seems to have a more prominent effect on the center and the west. Furtheremore, the price interdependence of contiguous districts in the north is seen to be more significant and stronger than the south.