Modelling office market dynamics: panel estimation and comparison of US metropolitan areas

A considerable body of research exists on how office rents and vacancy rates adjust in response to changes in demand and supply. In particular, recent research has settled upon an Error Correction Model (ECM) approach for investigating how adjustment occurs. Some recent studies have applied an ECM approach to panel data, but there remains only limited investigation of the cross sectional variation in such data and what factors drive differences in behaviour between different markets. For this reason, the present paper uses panel data for US office markets and analyses differences in the parameters found for different locations. The data consist of quarterly observations for rents, vacancy rates, employment, stock and completions for office markets in around 60 US MSAs. From this data, two samples are isolated: a set of 18 markets for which all variables are available from 1980 to 2011 and a set of 58 markets where they are available from 1988. Panel estimations are performed and these estimations are followed by further analyses that seek to explain the cross-sectional variation in results using variables that depict the characteristics of locations in terms of their economies and real estate markets.