Three essays on macroeconomics and applied econometrics

This dissertation consists of three essays investigating different economic phenomena with varying econometric techniques. Analyses in these essays used panel econometrics, spatial econometrics, and Factor Augmented Vector Autoregressions (FAVAR).;In the first essay Seemingly Unrelated Regression with Panel Data: Mixed Effects and Hausman-Taylor Estimators, we explore a seemingly unrelated regression (SUR) mixed effects estimator that allows for a combination of fixed and random effects, and a Hausman-Taylor estimator that addresses endogeneity among explanatory variables with instrumental variables. The main contribution of this paper is that we complete the theory of panel SURs and explore a method to deal with between-equation correlations unique to panels. We examine pretest strategies to select an estimator when the structure of the system is uncertain.;In the second essay The Bank Lending Channel: A FAVAR Analysis, a FAVAR is used to identify changes in bank lending that initiate U.S. business cycles (the Bank Lending Channel). The FAVAR technique exploits a large number of macroeconomic indicators to identify monetary policy shocks and allows us to jointly analyze the response of lending at the aggregate and individual bank level drawn from Call Reports submitted to the Chicago Fed since 1976. We find that the Bank Lending Channel is more prevalent than previously thought using aggregated lending data, but the lending response of individual banks is diffuse and largely insignificant. Individual banks are driven by specific innovations more than common monetary shocks.;In the third essay, we use Spatial Autoregressive (SAR) model to examine heterogeneity in the simultaneous space-time impact of foreclosures on neighborhood property values for foreclosed properties which experienced different types of foreclosure outcomes. We find that for incomplete foreclosures, the price impacts begin while the foreclosing household still has ownership of the property and continue throughout the foreclosure process. However, for complete foreclosures, the price impacts do not occur until at least 6 months after the foreclosing household has lost ownership of the property. Further, houses that sell within 6 months of the household losing ownership of the property are associated with very little neighborhood price externalities.