Answers to Which Questions? Clarifying Estimands to Rethink the Muslim–Antisemitism Link in Germany

Czymara et al. (2025) find that antisemitism is more prevalent among Muslims than among Christians and the religiously unaffiliated. They examine whether these relationships depend on educational attainment or religiosity and aim to explain them using a set of variables, including region of origin, religiosity, and education. I argue that adopting the estimands approach clarifies the research questions, reveals key identification assumptions, and demonstrates the need for causal mediation analysis. Using this approach, I distinguish between confounders and mediators, present revised estimates of what might be considered total effects, and examine the mediating and moderating roles of education and religiosity. The results indicate that caution is warranted when interpreting the link between Muslim denomination and antisemitism. Without controlling for confounders, total effects are overstated. Furthermore, differences in antisemitism between Muslims and Christians are minor among less religious people. However, even among highly educated Muslims, a substantial gap remains relative to Christians and the unaffiliated. I conclude by outlining how the estimands framework may help accomplish one key aim of Czymara et al. (2025): identifying mechanisms that could inform the design of targeted interventions to mitigate antisemitism.