Peer-reviewed Journal Articles
Klara Müller. 2025. “Survey Nonresponse After Elections: Investigating the Role of Winner-Loser Effects in Panel Attrition”, International Journal of Public Opinion Research 37(3): edaf031. https://doi.org/10.1093/ijpor/edaf031.
Abstract:
When and for whom do election outcomes drive survey nonresponse? This paper investigates whether belonging to the winners or losers of an election affects the likelihood of participating in post-election surveys. Systematic winner-loser effects on survey participation and resulting sample disproportionalities could yield decisive consequences for the accuracy of public opinion measurements. Through a series of logistic regressions and survival analyses on individual-level panel data from the German Longitudinal Election Study (GLES) after four German federal elections between 2009 and 2021, I find that winners and losers are equally prone to post-election survey nonresponse. Attitudes towards the democratic system and past survey loyalty outweigh winner-loser effects: For both winners and losers, more negative democratic views seem to make people more likely to exhibit (durable) post-election nonresponse. Moreover, past survey participation reduces post-election nonresponse and its durability. Crucially, my results show that election outcomes alone rarely change survey participation. However, post-election surveys tend to display underrepresentation (of, e.g., less educated or politically uninvolved people), if people who disapprove of the election outcome are already likely to have more cynical attitudes towards politics and the democratic system.
Lehrer, Roni, Oke Bahnsen, Klara Müller, Marcel Neunhoeffer, Thomas Gschwend, Sebastian Juhl. 2024. “Rallying around the leader in times of crises: The opposing effects of perceived threat and anxiety”, European Journal of Political Research 64: 697-718. https://doi.org/10.1111/1475-6765.12717.
Abstract:
In times of crisis, citizens tend to increase their approval of the government and its leader which can shift the balance of power. This ‘rally effect’ is a persistent empirical regularity; however, the literature is still undecided on its underlying causal mechanisms. We argue that crises induce threat and anxiety and hypothesize that perceived threat increases approval of the incumbent leader, whereas anxiety decreases it. By analysing German panel data from the COVID-19 pandemic, we causally identify both mechanisms and provide systematic evidence supporting this theory. Moreover, we increase the scope of our theory and show that both mechanisms are also at work when citizens approve cabinet members who manage key portfolios. Our findings have highly important implications for our understanding of the rally effect and crises politics in democracies.
Gschwend, Thomas, Klara Müller, Simon Munzert, Marcel Neunhoeffer, Lukas F. Stoetzer. 2022. “The Zweitstimme Model: A Dynamic Forecast of the 2021 German Federal Election“, PS: Political Science and Politics 55(1): 85-90. https://doi.org/10.1017/S1049096521000913.
Abstract:
In our election forecasting model project “Zweitstimme.org”, we developed a model to predict party-vote shares, coalition shares, the likelihood of a majority for certain coalitions, and many other relevant quantities of interest for German federal elections. In this article, we apply the model to the 2021 election.
Our point of departure is a Bayesian forecasting approach that combines polls and fundamentals. This follows the tradition of synthetic forecasting models (Graefe 2017; Lewis-Beck and Dassonneville 2015), which combine the merits of fundamentals-based with poll-based models. This article applies our dynamic Bayesian forecasting model to predict the outcome of the 2021 German federal election. It systematically combines published pre-election public-opinion poll results with information from fundamentals-based forecasting models while also accounting for the dynamic evolution of party support in multiparty systems.
We present an early forecast of our model, calibrate it on the basis of historical data, and report various quantities of interest, including the probabilities of a plurality of votes for a party, a majority of seats for certain coalitions in parliament, and the expected overall size of parliament due to the distribution of overhang (i.e., surplus) and compensatory seats. In its current form, the model generates forecasts for various points of interest beginning as early as 200 days before Election Day.
Book Chapters
Schoen, Harald und Klara Müller. 2024. “Allein auf den Wahlkampf kommt es an? Eine Analyse der zeitlichen Lagerung von Wieder- und Wechselwahlentscheidungen zwischen den Bundestagswahlen 2017 und 2021“. In: Schoen, H., Weßels, B. (eds.) Wahlen und Wähler. Springer VS, Wiesbaden. https://doi.org/10.1007/978-3-658-42694-1_23
Abstract:
In diesem Aufsatz untersuchen wir die Wechsel- und Wiederwahlentscheidungen, die der Verschiebung der parteipolitischen Kräfteverhältnisse von der Bundestagswahl 2017 zur Wahl 2021 zugrunde liegen. Die Ergebnisse auf der Basis von Befragungsdaten aus dem GLES-Panel zeigen, dass im Wahlkampf 2021 etliche Parteiwechsel vollzogen, aber auch revidiert wurden. Zugleich wurden nicht wenige Wechsel, etwa zugunsten der Grünen, bereits vor dem Wahlkampf – und bis zur Wahl 2021 andauernd – getroffen. Die zeitliche Lagerung von Wechsel- und Wiederwahlentscheidungen hängt mit politischem Interesse noch schwächer zusammen als mit Parteibindungen. Jedoch können partei- und kandidatenbezogene Einstellungsänderungen vor und in dem Wahlkampf Wechselentscheidungen statistisch erklären. Dies gelingt vergleichsweise gut, wenn Einstellungsänderungen und Wechselentscheidungen in derselben Periode betrachtet werden.
Klara Müller 