UC San Diego
Essays on the specification of New Keynesian dynamic stochastic general equilibrium model
- Author(s): Jung, Yong-Gook
- et al.
The New Keynesian dynamic stochastic general equilibrium model has become one of the standard approach to monetary policy analysis and macroeconomic forecasting. Therefore, many researchers are trying to come up with a better specification of the model. My dissertation goes in line with those efforts. The first chapter focuses on the specification of investment, the second on the firm's pricing, and the third labor market. The first chapter addresses the importance of rigorous investment specification. Dynamic stochastic general equilibrium (DSGE) models have a difficult time accounting for the slow response of investment spending to economic shocks that are generally found empirically. Four different specifications of investment dynamics are examined: (1) time-to-build as modeled by Kydland and Prescott (1982), (2) time-to-build as modeled by Casares (2006), (3) time- to-build as modeled by Wen (1998), and (4) investment adjustment costs as modeled by Christiano et al. (2005). Maximum likelihood estimation results indicate that the model with the investment lag specification of Casares (2006) fits the data significantly better than the other models. The second chapter examines the indexation schemes in Calvo-style sticky price model. While sticky price DSGE models with the dynamic backward indexation are widely used for monetary policy analysis, statistical plausibility of the indexation scheme has not yet been verified. The maximum likelihood estimation result does not find much support for the DSGE model with the backward indexation. The fit of the DSGE model with the dynamic backward indexation is no better than that with the static indexation. Also, the forecast of the model with the backward indexation is less accurate than that with the static indexation. The third chapter shows that when the workweek of capital and efforts are allowed to vary, the employment lag itself cannot generate a hump-shaped response of output to a monetary shock. The reason for this is that despite the fact that the size of employment is predetermined, firms can rely on other low adjustment cost margins such as the workweek of capital and efforts to meet the increase in demand due to a positive monetary shock