Abstract

This thesis investigates three topics in theoretical and applied econometrics: Bartlett-type correction of the Distance Metric (DM) test, a Generalized Method of Moments (GMM) study of the effect of the North American Free Trade Agreement (NAFTA) on Quebec manufacturing industries, and a goodness-of-fit test for copulas. The first topic derives an Edgeworth approximation of the distribution of the DM test statistic and obtains a Bartlett-type correction factor, then it uses examples of covariance structures to illustrate the theoretical results and applies the theoretical results to study the covariance structure of earnings. The second topic calculates Canadian tariff rates over the period 1991-2007 for manufacturing industries, classified using the North American Industry Classification System (NAICS), proposes a simulation-based moment selection procedure to improve the properties of the system GMM estimator, and analyzes the effect of NAFTA on earnings of Quebec manufacturing industries. The third topic proposes a new rank-based goodness-of-fit test for copulas, conducts a power study to show that the new test has reasonable properties, and presents an application