UC San Diego

Motivated by the frictions found in the theoretical literature on the credit channel of monetary policy, chapter one investigates whether credit default swaps' (CDS) superior measurement of credit risk can be used to forecast real economic activity. In a simple model, we find that CDS improve forecasts when compared to standard measures of credit risk and portfolios of corporate bonds compiled by Gilchrist et al. (2010), particularly in the middle of the credit distribution. However the CDS data is limited by only having a short time series. Two additional models are then estimated to overcome this hurdle, a dynamic factor model and Bayesian model averaging (BMA). In these more robust settings, CDS lose their explanatory power. This suggests that modern dimension reduction techniques can successfully extract a parsimonious forecasting model in a data rich environment. The recent global financial crisis suggests the post-1984 Great Moderation has come to an abrupt end. How we obtained nearly 25 years of stability and why it ended are ongoing puzzles. Chapter two departs from traditional monetary policy explanations and considers two empirical regularities in US employment : i) the decline in the procyclicality of labor productivity with respect to output and labor input and ii) the increase in the volatility of labor input relative to output. We first consider whether these stylized facts are robust to statistical methodology. We find that the widely reported decline in the procyclicality of labor productivity with respect to output is fragile. Using a new international data set on total hours constructed by Ohaninan and Raffo (2011) we then consider whether these moments are stylized facts of the global Great Moderation. We document significant international heterogeneity. We then investigate whether the role of labor market frictions in the US as found in Galí and van Rens (2010) can explain the international results. We conclude that their stylized model does not appear to account for the differences with the US experience and suggest a direction for future research. Chapter three examines state-level dynamics of revenues and expenditures. In contrast to previous literature that consider budget deficits in a panel setting, we consider the income elasticity of revenues and expenditures in the cross-section. We examine three budget levels : 1) total budget, 2) excluding liquor stores, utilities, and insurance trusts (LUSI) revenue and expenditures, and 3) excluding capital expenditures. We find that every 1% increase in revenue elasticity is associated with a 0.57% increase in expenditure elasticity for the non-LUSI budget, and this is robust to alternative specifications