Essays in capital utilization

This dissertation has three chapters unified by a central theme: allowing for variations in capital utilization changes the dynamics of macroeconomic models. In Chapter 2, I explore a modification of Khan and Thomas (2007)'s (s,S) inventory model. Khan and Thomas (2007) present an inventory model suggesting that adjustment costs may help explain inventory movements over the business cycle, even in absence of sticky prices and with shocks arising only from productivity. By adding variable capital utilization to their model, the dynamics change and the results improve. In their original model, intermediate goods producers could only change their production in the short run by changing the quantity of labor input. This structure prevented intermediate goods production from being as variable as in the data, as labor was the only resource that was variable in the short run. When variable capital utilization is introduced, the supply of intermediate goods is more elastic. This chapter finds that variable utilization can change business cycle moments by changing the interactions between sectors in a macroeconomic model.;Chapter 3 evaluates one possible explanation for variations in utilization: cyclical changes in shift work. The empirical literature suggests that procyclical productivity can be explained by changes in the utilization of capital. Some suggest that this change in capital utilization is likely due to changes in shift work. Chapter 3 evaluates this claim in a DSGE model that explicitly accounts for shift work. I find that in a model with variable shift work, productivity is procyclical even when shocks are to preferences. However, the volatility in apparent productivity from the model is significantly less than in the data, and some business cycle moments have the wrong sign, except when the cost of running a second shift is in a specific form and the shocks have a specific structure. In contrast, a model that assumes that shift work is not variable does quite well matching business cycle moments - as long as productivity shocks are the driving force behind the business cycle. In short, variations in shift work seem like an unlikely explanation for observed procyclical productivity.;Chapter 4 uses Bayesian estimation techniques to estimate various versions of the DSGE model including shift work. The estimation establishes that a final goods cost for running a second shift is the version that most closely matches the data, that the inclusion of shift work diminishes the importance of productivity shocks in explaining macroeconomic fluctuations, and that productivity shocks are still important in explaining uctuations in output even when shift work choice is included in the model.;In summary, capital utilization matters. In multi-sector models, variable utilization provides an additional degree of flexibility that can allow for models to more closely match key business cycle moments. Simple macroeconomic models including variable capital utilization explain some of the appearance of productivity-driven business cycles, and also more closely match the data than models without variable shiftwork.