Maximum Likelihood Estimation Of Fixed Effects Dynamic Panel Data Models Covering Short Time Periods PDF Download

Chapter 1. Hansen (1999) introduced threshold estimation in non-dynamic panel threshold models. In this chapter we extend this previous work to allow dynamics in a panel data threshold model with individual fixed specific effects covering short time periods. We propose a maximum likelihood approach to estimate the structural parameters using a first difference transformation of a Dynamic Panel Threshold Model. We show the Maximum Likelihood estimation of the threshold parameter is consistent and converges to a double-sided standard Brownian motion distribution as in Hansen (2000), when the number of individuals grows to infinity for a fixed time period; and the Maximum Likelihood estimation of the slope parameters are consistent and converge to a normal distribution. Chapter 2. The super-neutrality of money hypothesis states that nominal variables do not affect real variables in the long-run. Nonetheless, Fischer (1993) found a negative relationship between inflation and economic growth, but Bruno and Easterly (1998) suggest that relationship is only present with high inflation periods. In this chapter we estimate a threshold level of inflation, above which inflation significantly slows growth; we estimate a dynamic panel threshold model. Using a sample of 72 countries and 8 periods of 5-year averages from 1961 to 2000, we found a threshold level of inflation at 13 percent, where inflation above this threshold has a negative effect on economic growth. In a model with a double threshold, we found two threshold levels of inflation at 13 and 39-42 percent, where that negative effect is stronger for inflation above 39-42 percent.