ASSESSING THE ACCURACY OF STRUCTURAL PARAMETER ESTIMATES IN ANALYSES OF IMPLICIT MARKETS
Academic Article
Overview
Identity
Additional Document Info
Other
View All
Overview
abstract
For years economists have attempted to estimate the demand for heterogeneous goods such as housing. In recent years, Rosen's (1974) model of implicit markets has become generally accepted as the appropriate vehicle to analyze a heterogeneous good empirically. The sensitivity of three-stage least-squares (3SLS) estimates of structural parameters to changes in the choice of normalization may be used as an ad hoc indicator of their accuracy. Accuracy is related to the degree of "effective' identification in the system. Using a hypothetical model, the similarity of structural estimates under alternative normalizations is compared to the true (unobservable) accuracy of the estimates. Estimates that are relatively insensitive to normalization are more accurate in general than estimates that are sensitive. The implications of the results for practical empirical applications are then discussed. -from Authors
