Determinants of Geographic Unemployment Rates: A Selectively Pooled-Simultaneous Model

Using a procedure that allows selective pooling of certain subsamples of the data, this study examines the determinants of state unemployment rates in the United States. By selectively pooling the data, more reliable parameter estimates result than could otherwise be obtained were the data disaggregated to the state level, and the specification error inherent in a complete pooling of the data is avoided. The models are estimated using two stage least squares, and the empirical results indicate considerable heterogeneity across areas in the ways in which unemployment rates are determined.