Polarization as a Result of Populism? Evidence from Plenary Debates in the Bundestag

For decades, Germany had seemed to withstand a populist radical right upsurge. Yet this has changed with the rise of the Alternative for Germany (AfD) from 2013 onwards and its arrival in the Bundestag in 2017. The AfD articulates hostile positions towards established politics and elites as well as radical right positions directed against migrants, refugees, and minorities. Elsewhere, we investigated whether the presence of the AfD in the Bundestag has led to a decline in its core parliamentary functions (Schwanholz, Lewandowsky, Blätte, Leonhardt 2020, forthcoming). In a broader perspective, contemporary democracies are characterized by increasing polari-zation (e.g. Levitsky and Ziblatt 2018). More specifically, parliament can be considered as the ‘core institution’ in representative democracies where important political debates between the parties take place. It is thus crucial to examine the impact of the populist radical right on the communicative function of parliament. We therefore analyze whether the representation of the AfD in parliament has resulted in a greater division of the parliamentary discourse. Our contribution will further inspect parliamentary communication not only with regard to the AfD but also in comparison to discursive characteristics before its emergence. In short, we are thus able analyze whether, first, polarization takes place and, secondly, if it is a result of the presence of a populist radical right party. To measure polarization, we rely on an extensive corpus of parliamentary speeches. We also move beyond text-based approaches that try to estimate ideal points, such as Wordscore, Wordfish or Wordshoal and employ a machine learning approach that focuses on differences in the language that is used by specific actors (Blätte 2018; Goet 2019). This provides the capacity to estimate the level of polarization over a long period of time while retaining the abil-ity to gain an insight in the disputed objects as manifested in parliamentary language. The data we use is an extended version of the GermaParl corpus available for 1990 – 2018 at this stage, which will be extended to cover the 1970s and 1980s for the purpose of this paper. We can thus contribute to a better understanding of the impact of populism and its consequences over a period of more than 40 years – also with regard to modern-day populism of the AfD.