Dynamic limited dependent variable modeling and United States monetary policy

My dissertation is motivated by the observation that there are several contexts in applied macroeconomic research where the support space of the dependent variable being studied is subject to substantial restrictions of various forms. For instance, the dependent variable can be discrete, or it can be censored. Such contexts call for Limited Dependent Variable (LDV) modeling, but they also typically call for dynamic modeling.However, most of the existing empirical work either ignores the restrictions to the support space of the dependent variable under consideration and focuses on the needed time series modeling requirements, or, conversely, employs LDV estimation techniques at the expense of time series modeling. The main reason for that is that estimating LDV Time Series models through standard extremum methods can be a daunting computational task because of the need for integration of high order multiple integrals and/or numerical optimization of difficult objective functions.In this dissertation I propose a classical Markov Chain Monte Carlo estimation approach with data augmentation that overcomes both of these problems and I apply this approach to two areas in U.S. monetary policy:In the first chapter I establish the asymptotic properties of the proposed estimator and I also propose a practical and flexible algorithmic framework for this class of models, and I illustrate its performance in small samples using simulated data.In the second chapter I estimate, using real-time data, a forward-looking, dynamic, discrete-choice monetary policy reaction function for the U.S. economy. I find a substantial contrast between the periods before and after Paul Volcker's appointment as Fed Chairman in 1979, both in terms of the Fed's response to expected inflation and in terms of its response to the (perceived) output gap: In the pre-Volcker era the Fed's response to inflation was substantially weaker than in the Volcker-Greenspan era conversely, the Fed seems to have been more responsive to real activity in the pre-Volcker era than later.In the third chapter I estimate a dynamic Tobit model with Time-Varying parameters for the daily reaction function of the Open Market Desk of the U.S. Federal Reserve. The results reveal a rich pattern of dynamic behavior by the Open Market Desk both inside the maintenance period and across maintenance periods and point towards a Desk which is highly adaptable to evolving conditions both in the economy in general and in the market for reserves in particular.