Business-Administration - Course - International Trade and Trade Policy

This course will emphasize models that are not fully specified (parametrically). For example in estimating the conditional mean ~for a random variable of interest, we will not assume that its functional form (e.g. linear in the case of the linear regression model) is known. Alternatively, in estimating a discrete choice model for labor force participation, parametric methods commonly assume that the value (indirect utility) to joining the labor force has a known distribution (e.g. normal in the case of the probit model). Here, we will examine distribution-free methods rather than making any parametric distributional assumption.