Seminar: Current Research in Voting and Democratic Institutions - Examining the Replicability of Selected Papers in Political Economy

Course description

Transparency and reproducibility are topics of growing interest in economic research. The heavily discussed „replication crisis“ in economics, psychology and other disciplines emerged when a considerable fraction of empirical studies (experimental and observational) could not be reproduced. A famous (now infamous) example is Reinhart/ Rogoff (2010). Improving the standards of empirical economic research, many journals now require pre- analysis plans and open sharing of data and materials.

The aim of this seminar is to critically reflect the value and credibility of research in the field of political economy. Students learn how to embed papers in the overall scientific literature and replicate (parts of) them using the statistical software STATA. We will clarify which contribution the respective paper makes. Is the identification strategy credible? Which method was implemented? How suitable is the used dataset? What are the advantages and disadvantages of the respective approach in order to answer the research question at hand? The skills acquired will be useful in the context of writing the master thesis.

In the seminar, students write a short research paper, in which they replicate or extend an existing paper in the field of political economy. They will present the main features of the paper and their results. Articles including dataset and STATA do-files will be suggested, but students are free to propose their own paper that they wish to replicate, subject to approval. Since the focus of this seminar is empirical, previous knowledge of STATA is helpful, but not a requirement. We will introduce the software STATA in a tutorial session.

The extension can focus on a subsample, include additional explanatory variables, employ an alternative methodology or use additional data. The extension may provide additional intuition or investigate the robustness of the results in the original paper.