This paper aims to examine the housing market in Greece after the Global Financial Crisis focussing on regional analysis and urban markets in Athens and Thessaloniki.Design/methodology/approachThe paper uses a data set of over 70,750 property values from 2007 until 2014 incorporating characteristics variables upon which hedonic models are estimated. These form the bases for calculating value indices for mix adjusted houses/apartments by year and region. The indices are used in a panel model in which regional and economic variables are included as independent variables. Using advances in dynamic panel data modelling, a bias-corrected least squares dummy variable corrected (LSDVC) model is applied.FindingsResults indicate the importance of macroeconomic variables in terms of the role of disposable income and significantly different regional effects. Examining the major urban markets, results indicate significant differences in the response of house values to exogenous demand side influences, consistent with the finding of significant regional differences in the LSDVC.Research limitations/implicationsWhile data on valuations are used that may contain smoothing, the data set covers a large sample of residential properties. As regional economic differences are significant and persistent, housing markets will also behave differently, and hence national policies, unless targeted, will have regionally differentiated effects.Practical implicationsRegional heterogeneity needs to be considered in model estimation.Social implicationsPolicymakers should consider regional differences to improve policy effectiveness.Originality/valueThis is the first paper to use a large sample of residential properties in Greece and apply the LSDVC model to overcome estimation biases.
Journal of European Real Estate Research – Emerald Publishing
Published: Nov 21, 2018
Keywords: Greece; Housing; Regional; Panel estimation