The Impact of Activity in Instagram on the Cost of Housing on the Secondary Market

Property is an element of national wealth so identification of basic determinants of housing prices is a purpose of study for researchers. In literature the influence of various characteristics such as structural and neighborhood is explored by using different estimating techniques. One of the most common is hedonic price method based on OLS estimation and some methods which take into account spatial effects and spatial heterogeneity like geographically-weighted regression. Recent studies also use data from social networks in housing prices analysis. In this paper we use data from EMLS 24 and Instagram in order to explain housing sales prices in St.Petersburg. We complement housing dataset with new variables reflecting the activity in various places of the city by calculating Getis-ord G_i^* statistics and apply two estimation approaches: standard hedonic price method and geographically-weighted regression in order to see which characteristics influence sale prices in St.Petersburs and how spatial effects can be taken into account. Results of the estimation reveal that housing prices are affected by both structural characteristics and degrees of activity in different locations in St.Petersburg and the coefficients vary in space. Moreover, this can potentially coerce more researchers to use data from social media in this scientific field.