An examination of the bank lending channel from a regional perspective

This dissertation intends to investigate the existence of the bank lending channel from a regional perspective. To realize this objective, it builds a theoretical model and empirically tests the predictions derived from the model. It shows that the importance of the bank lending channel declines as bank size and/or firm size increase. Based on this result and the observation that the compositions of banks and firms vary substantially across the nine census regions in the United States, it predicts that if the bank lending channel is operating, changes in monetary policy have cross-regional differential effects on bank balance sheets and real economic activity.; To test the theoretical predictions about the bank lending channel, this dissertation first selects the Middle Atlantic region to represent regions with a small concentration of small banks and small firms and the West North Central region to represent regions with a large concentration of small banks and small firms. It then compares the behavior of bank balance sheets and real economic activity between these two regions by conducting two types of analysis—a static analysis and a dynamic analysis—with a VAR model.; This dissertation observes mixed results in the static analysis, which includes a series of Granger-causality tests and comparisons of coefficient estimates. However, it finds more favorable results for the existence of the bank lending channel in the dynamic analysis, which includes decomposing forecast error variances and computing impulse response functions. Based on the overall results of these analyses, it asserts that there is evidence in support of the bank lending channel. This is especially so if the focus is on the results of the dynamic analysis and an assumption of heterogeneous demand for bank loans is adopted. Nonetheless, it also points out that the evidence is weak in the sense that it is not robust to the lag length of the VAR model and the position of monetary policy indicator among the variables. Therefore, this dissertation concludes that the existence of the bank lending channel cannot be unambiguously established.