Use of Hedonic Indices to Distinquish Real Changes in Housing from Changes in Housing Expenditures: Evidence from the Housing Allowance Demand Experiment. (RE-ANNOUNCEMENT of PB85-127066-see notes field for explanation)

摘要

This paper discusses the use of hedonic indices to allocate changes in housing expenditures between real changes in housing and price changes due to inflation; changes in conditions of tenure; and in particular, more or less effective shopping by either individual households or government agencies. The study focuses on both behavioral and methodological issues. In terms of behavior, evidence indicates that the cost of any given type of housing unit varies within a given housing market, that households achieve better or worse deals through more or less careful shopping as well as by chance, and that household shopping behavior responds to changes in shopping incentives. Because of this, estimates of changes in housing expenditures may produce substantial error in estimating the real changes in housing associated with changes in housing prices or with participation in various housing programs. Evidence for these conclusions is based on comparison of household housing expenditures with the estimated hedonic index value of the housing services actually obtained; this is, in effect, an analysis of the hedonic residual. Such analysis may be biased by omitted variables in the hedonic index. The paper's methodological focus is on testing for and quantifying the effects of omitted variables in the estimated index. The study analyzed the Housing Allowance Demand Experiment and is framed in terms of renters. Similar considerations would apply to analyses involving the sale as opposed to rental of housing units. Appended are the hedonic equations and descriptions of the exact conditions for the use of income to identify omitted variable effects for other expenditure changes. Tabular data and 44 notes are provided.