Three essays on commercial indices

The intent of this dissertation is to advance the research in commercial markets by empirically examining three important issues of commercial real estate indices. The empirical investigation of these issues is presented in three separate essays. The first essay empirically examines office rent determinants during market decline and recovery by using results from a constant-quality hedonic index generated for the Phoenix Metropolitan Area. The parameters from the hedonic technique estimate the implicit prices of the determinants of office rents. The examination finds that office rent determinants vary between periods of market decline and recovery. The study also examines the dependent variable specification of the hedonic rent determinant model.; The second essay investigates sample selection bias in commercial indices. Real estate price indices may suffer from selectivity bias because samples of sold properties, used to construct the index, may not accurately represent the population of properties. A unique data set in the Phoenix area allows for an examination of the sample compositions of transacting and non-transacting properties and allows for an explicit investigation of sample selection bias. The analysis finds that selectivity bias exits in the sample of sold commercial office properties. Rental and price indices corrected for sample selection bias provide insight into the Phoenix office market.; The third essay examines time-varying parameter techniques for constructing commercial indices, corrected for sample selection bias. The conventional hedonic technique for constructing indices holds constant over time the implicit prices of the quality characteristics. If the implicit prices vary over time, this process introduces another form of bias into the index. This third study empirically investigates two different time-varying parameter techniques to overcome this problem. Methods for examining and correcting sample selection bias are also incorporated into these time-varying techniques. The empirical examination finds that time-varying parameter techniques improve upon the conventional technique for constructing constant-quality commercial indices. The analysis also finds and corrects for sample selection bias in the varying parameter indices.