Current Research

Political scientists pervasively use data that contains sensitive information – e.g. micro-level data about individuals. However, researchers face a dilemma: while data has to be publicly available to make research reproducible, information about individuals needs to be protected. Synthetic copies of original data can address this concern, because ideally they contain all relevant statistical characteristics without disclosing private information. But generating synthetic data that captures–eventually undiscovered–statistical relationships is challenging. Moreover, it so far remains unsolved to fully control the amount of information disclosed during this process. To that end differentially private generative adversarial networks (DP-GANs) have been proposed in the (computer science) literature. We experimentally evaluate the trade-off between data utility and privacy protection in a simulation study by looking at evaluation metrics that are important for social scientists, specifically in terms of regression coefficients, marginal distributions and correlation structures. Our findings suggest that on average, higher levels of provided privacy negatively affects the synthetic data quality. We hope to encourage inter-disciplinary work between computer scientists and social scientists to develop more powerful DP-GANs in the future.
Working Paper, 2019

In this paper, I seek to develop a measurement for vague language in written constitutional court rulings. I use two different methods to approach this: a dictionary approach expanded using word embeddings, and a machine learning classifier (using both traditional NLP classifiers and recent deep learning classifiers).
Working Paper, 2019

In this dissertation project, I investigate how constitutional courts use vague language to strategically either apply pressure to the government or hide likely legislative non-compliance from public view. I find that in line with a popular game-theoretic model (Staton and Vanberg (2008)), courts use vagueness strategically depending on their level of public support and policy-uncertainty.
Working Paper, 2018

This study uses random forests to predict over 2,900 proceedings decided by the German Federal Constitutional Court between 1974 and 2010 using only information available prior to the respective decision with a machine learning classifier. In particular, I test the importance of legal context and/or political context factors for the prediction. My results show that legal context alone is already a good predictor for the outcome of a case, but that the predictive performance can be significantly improved when information about the political context of a decision is added.
Working Paper, 2018

How do citizens evaluate candidates for highest courts? This paper employs a discrete-choice experiment in order to untangle the relative importance of both dimensions for gaining public support of a judicial nominee and to identify the type of nominee the public prefers most. Our results clarify the conditions under which a judicial candidate’s perceived lack of political independence can be compensated.
Under Review, 2018

In this paper, we use a quasi-natural experiment to study how party label heuristics influence the replacement policy of absentee judges at the German Federal Constitutional Court.
Working Paper, 2016

Upcoming Talks

I present a current Working Paper of mine called: How to Forecast Constitutional Court Decisions? Legal and Political Context in a Machine Learning Application Here, I investigate whether it is possible to forecast decisions of the German Federal Constitutional Court using machine learning.

Teaching

I am a teaching instructor for the following courses at University of Mannheim:

Besides that, I am an instructor of professional training projects:

  • November 2017: Introduction to Supervised and Unsupervised Machine Learning, 3 days workshop, Deutsche Bundesbank, Frankfurt
  • February 2018: Introduction to R, 1 day workshop, Geschäftsstelle für Qualitätssicherung Hessen, Frankfurt
  • April 2019: Big Data Analysis and Introduction to Supervised and Unsupervised Machine Learning/Deep Learning, 5 days workshop, Deutsche Bundesbank, Frankfurt

Testimonials:

Awesome tutor, probably the best course I had so far combined with the lecture. Really great job, always explained everything with calm in non-technical terms that also non-statisticans understand.


Maybe the most motivating learning environment I have yet experienced!


This was actually one of the best and most interesting courses I ever had during my university career! I really liked attending the course and learned a lot. The lecturer was very friendly, patient and always helpful. During the course, his office hours and even via email he was always available and helped to solve even very complicated topics. That made the quite difficult and sometimes very exhausting weekly problem sets bearable. I got the impression that the lecturer was very motivated and always well prepared. Well done!


Really passionate lecturer! Good job!


Contact