Workshop report „Causality in the Social Sciences II“

Lang, Sebastian | Schwabe, Ulrike | Jungbauer-Gans, Monika

Based on our amazing experiences attending the workshop on “Causal Inference in the Social Sciences” at the University of Leipzig in summer 2019, we decided to host a follow-up workshop at the German Centre for Higher Education Research and Science Studies (DZHW) in Hanover. The subsequent workshop “Causality in the Social Sciences II” has been integrated into the scientific activities of DZHW’s research cluster “Empirical Methods in Higher Education and Science Studies”. Causal analysis is one thematic focus among others. The particular objective of this workshop was to bring together young scholars from different fields to discuss all questions related to causality. As already announced in the Call for Paper, the workshop aimed to discuss and reflect up-to-date developments in research and open questions on causality from an interdisciplinary perspective and focussed on three purposes:

1. To deepen the understanding of causality in the social sciences on a theoretical-analytical level.
2. To highlight research designs and statistical methods to support causal interpretation on an applied empirical level. This implies experimental research designs as well as estimation strategies for observational data, and the critical reflection of strengths and pitfalls.
3. To provide a place to discuss results of replication studies directly confronting naïve and more sophisticated estimation strategies: Do different methods provide different results?

Hence, core questions to be addressed were related to the general relation between description and explanation, the counterfactual approach itself, adequate estimation strategies for causal inference, advantages and disadvantages of experimental and observational data.
We were overwhelmed by the number and the high-quality standards of the submitted abstracts and have choosen 23 presentations. We grouped them according to theoretical-analytical contributions (e.g. DAGs), the particular methods applied (e.g. experimental settings, panel regression) and areas of interest (e.g. health, education, labour market). Besides the research paper presenters, we were able to invite three out-standing, international researchers from the field of economics and sociology to give keynotes:

• Prof. Dr. Jan Marcus (University Hamburg, Germany): “Federalism and causal inference: Chances and challenges of difference-in-differences”.
• Prof. Dr. Felix Elwert (University of Wisconsin-Madison, USA): “Mehr Experimente wagen: Causal Inference for Higher Education Research”.
• Prof. Dr. Michael Lechner (University of St. Gallen, Switzerland): “Priority to unemployed immigrants? A causal machine learning evaluation of training in Belgium”.

Originally, we planned a two-day intensive workshop with a maximum number of 30 participants. But in times of pandemic, we had to change this plan in August 2020 and switched to a fully online format. As a consequence of switching the format, we extended the workshop to three days and opened it for a broader interested audience. We were overwhelmed a second time by the great interest in the workshop within the scientific community. In total, there were 23 active (“presenting”) participants and the participants came from various countries (Italy, Russia, the Netherlands, and Germany) as well as disciplines (Philosophy, Economics and Sociology). In addition, 132 passive (“listening and discussing”) participants registered themselves via our website. Altogether, the participants presented 57 different universities and 17 research institutes. While the majority were Germans, 45 international participants attended the online-workshop after all.
Wrapping up these three workshop days, the participants concluded that it was a good idea to switch to online as travelling to Hanover would have been very difficult to manage for most of them. Many of them confirmed that they received valuable feedback on their presentations, which will help them to further improve their presented work. The chat function within the software “WebEx” was used to inform others about relevant literature. Additionally, breaks and special breakout rooms were used for in depth discussion among smaller groups of participants. With respect to the presented content, we conclude first that philosophers can substantially contribute to the discussion on evidence-based policy on a meta-level. In general, the interdisciplinary of the participants was second extremely fruitful and disciplinary boundaries seem to rather hold back scientific progress. Third, the “gold-standard” of doing experimental research was critically discussed under the perspective of external validity of the findings. And finally, we noticed that causal thinking is disseminating into the community of higher education research as well.

Finally, we would like to thank the Academy of Sociology for this outstanding opportunity to host the Workshop “Causality in the Social Sciences II” at the DZHW. At the end of the 2020 workshop, we suggested to continue with the idea of a continuous exchange on causal questions by organizing another follow-up workshop “Causality in the Social Sciences III” among the participants. This would lead to some kind of institutionalization of causal thinking in sociology, especially among young scholars. We would highly appreciate if the Academy of Sociology would share our opinion on that recommendation.