The course consists of two parts. The first part consists of more advanced theory and methods relating to causal approaches surpassing the multivariate linear regression, limited dependent variable regression and time series analysis covered by Econometrics I. It also considers how to apply these methods through examples of how such methods are used in economic history. It discusses issues like selection bias, the bad control problem, and unobserved heterogeneity and the pitfalls associated with them as well as the possibilities to deal with these issues. This part advances the knowledge of empirical analysis making use of computer software (e.g. Stata). In the second part of the course, students independently analyse a more advanced quantitative problem using actual data from economic history, and report results in individual papers, showing awareness of the pros and cons of various causal approaches in econometrics.