HEDONIC MODELING OF OFFICE RENTS WITH SEMIPARAMETRIC GEOADDITIVE APPROACH: A CASE OF ISTANBUL OFFICE MARKET

Abstract

This study aims to establish a hedonic model for office space rents in Istanbul using a dataset of 2348 office spaces. The dataset includes information on office space and building characteristics, as well as lease, locational, and neighborhood characteristics that influence office rent. To analyze the relationship between office rent and these determinants, we utilized a semiparametric geoadditive model that allows for functional form flexibility and captures locational effects through geographical smoothing functions instead of locational dummies. Our findings suggest that the relationship between office rent determinants and office space rents is nonlinear in the nonparametric part of the model. Among the office rent determinants, the vacancy rate - a proxy of office space demand - has the greatest impact on office rents. The second most effective determinant is Class A, an indicator of office building quality, and the third most effective determinant is the possession of a Bosporus view.