Teaching

Econometrics

April 2017 Standard errors and serial correlation in panel studies [slides][dofile 1, 2]

I show where standard errors come from and what the difference is between regular SE, robust SE and clustered SE, both mathematically and intuitively. The second part takes this knowledge and illustrates how to apply this to day-to-day research issues, highlighting when each method is preferable. The third and final part looks at the issue of serial correlation and how to deal with it - e.g. adapting the Hausman test as well as introducing a barrage of alternative specifications that may be more credible if the standard methods show serially correlated errors.Guest lecture in the course, "Advanced Applied Econometrics" at KU Leuven (2017).

January 2018 Synthetic control methods [slides][dofile]

Synthetic control methods are popular, but an applied guide to using them in practice remains hard to find. In this presentation, I highlight the strengths and weaknesses of this method, explain the machinery behind the synth command in Stata and compare it to different treatment estimators.

May 2018 Treatment effects estimators [slides][dofile]

Treatment effects estimators are getting more and more popular and although there is ample guidance on how to use particular methods, it's harder to find a more general overview. I focus more on the motivation to do difference in differences (DiD), matching (nearest neighbour/propensity score) and synthetic control methods (SCM) and when you can expect one to work better than the other. The dofile provides a very hands-on demonstration on how to apply all of these in Stata. Special attention is devoted to the assumptions behind the various methods, especially the what the common trend assumption is and what it is not.Guest lecture in the course, "Advanced Applied Econometrics" at KU Leuven (2018).

Minimum Wages

April 2021 The Peculiarities of the Minimum Wage [slides]

Minimum wage policy challenges the classic ECON 101 thinking. In this presentation, I discuss briefly why a simple supply-demand competitive labour market does not work well in the minimum wage content. I conclude by discussing the redistribution effects of the minimum wage and what that implies for the optimal minimum wage level. Intuition only, no graphs nor equations.Guest lecture in the course, "Labour Economics" at KU Leuven (2021).