The EPINetz Twitter Politicians Dataset 2021. A New Resource for the Study of the German Twittersphere and Its Application for the 2021 Federal Elections

König, Tim; Schünemann, Wolf J.; Brand, Alexander; Freyberg, Julian; Gertz, Michael Article — Published Version The EPINetz Twitter Politicians Dataset 2021. ANew Resource for the Study of the German Twittersphere and Its Application for the 2021 Federal Elections Politische Vierteljahresschrift Provided in Cooperation with: Springer Nature Suggested Citation: König, Tim; Schünemann, Wolf J.; Brand, Alexander; Freyberg, Julian; Gertz, Michael (2022) : The EPINetz Twitter Politicians Dataset 2021. ANew Resource for the Study of the German Twittersphere and Its Application for the 2021 Federal Elections, Politische Vierteljahresschrift, ISSN 1862-2860, Springer Fachmedien Wiesbaden GmbH, Wiesbaden, Vol. 63, Iss. 3, pp. 529-547, https://doi.org/10.1007/s11615-022-00405-7 This Version is available at: https://hdl.handle.net/10419/306385 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. https://creativecommons.org/licenses/by/4.0/ RESEARCH NOTE https://doi.org/10.1007/s11615-022-00405-7 Polit Vierteljahresschr (2022) 63:529–547 The EPINetz Twitter Politicians Dataset 2021. A New Resource for the Study of the German Twittersphere and Its Application for the 2021 Federal Elections Tim König · Wolf J. Schünemann · Alexander Brand · Julian Freyberg · Michael Gertz Received: 2 March 2022 / Revised: 19 May 2022 / Accepted: 19 May 2022 / Published online: 15 June 2022 © The Author(s) 2022 Abstract This research note introduces the EPINetz Twitter Politicians Dataset, a comprehensive dataset of 2449 Twitter accounts of German parliamentarians, minsters, state secretaries, parties, and ministries on a state, federal, and European Union level for the year 2021. This hand-curated dataset not only provides up- to-date information on elected officials, but it also includes additional variables such as their party affiliation, age, and gender. Furthermore, it provides linkages to additional data sources by providing the accounts’ Wikidata and Abgeordnetenwatch (Parliamentwatch) IDs. While it does not provide actual tweet data, the dataset will be a valuable resource for researchers by providing easy access to elected German politicians. We demonstrate some of the dataset’s uses with an analysis of the 2021 German Federal Elections. The full dataset can be accessed via https://doi.org/10. 7802/2415. Keywords Social Media · Germany · Data · Political communication · Political representatives · Election campaigns Tim König () · Wolf J. Schünemann · Alexander Brand Institute of Social Sciences, University of Hildesheim, Hildesheim, Germany E-Mail: [email protected] Julian Freyberg · Michael Gertz Institute of Computer Science, Heidelberg University, Heidelberg, Germany K 530 T. König et al. Der EPINetz-Twitter-Politicians-Datensatz 2021. Eine neue Ressource für die Erforschung der deutschen Twittersphäre und ihre Anwendung auf die Bundestagswahl 2021 Zusammenfassung Die Research Note stellt den EPINetz–Twitter-Politicians-Da- tensatz vor, eine umfassende und aktuelle Sammlung der Twitter-Accounts deutscher Parlamentarier:innen, Minister:innen, Staatssekretär:innen, Parteien und Ministerien auf Länder-, Bundes- und europäischer Ebene für das Jahr 2021. Die insgesamt 2449 händisch kuratierten Accounts sind darüber hinaus mit Variablen wie Parteizugehörigkeit, Alter und Geschlecht annotiert. Zusätzlich enthalten sie die zugehörigen Wikidata- und Abgeordnetenwatch-IDs, um Verbindungen des Datensatzes mit diesen Datenquellen zu vereinfachen. Der Datensatz selbst stellt keine Twitterdaten zur Verfügung, sondern soll es Forschenden ermöglichen, einfach auf hochwertige Daten zu den Twitteraktivitäten politischer Repräsentant:innen in Deutschland zuzugreifen. Wir demonstrieren einige mögliche Anwendungsszenarien für den Datensatz mit einer Analyse der Bundestagswahlen 2021. Der vollständige Datensatz steht unter https://doi.org/10.7802/2415 zur Verfügung. Schlüsselwörter Soziale Medien · Deutschland · Daten · Politische Kommunikation · Politische Repräsentanten · Wahlkampf 1 Introduction This research note presents the EPINetz Project’s Twitter Politicians Dataset, a comprehensive overview of institutionalised German politics on Twitter. Over the years, digital media have been recognised as playing an increasingly relevant role in political communication. By now, it has become a widely accepted truth that social media have taken their place among the arbiters of public opinion (Margetts 2019; Jungherr et al. 2020). While assessments of their actual impact on democracy vary (Persily and Tucker 2020), social media have become a cornerstone for politicians’ efforts at campaigning, agenda-setting, and communicating with their constituencies (Bossetta 2018; Dimitrova and Matthes 2018). This can in no small part be attributed to the complex intermedia effects between social and “legacy” media (Jungherr et al. 2019a; Su and Borah 2019; Langer and Gruber 2021). The platform Twitter, in particular, has gained scholarly attention due to its high popularity among politicians and journalists (Dagoula 2019; Molyneux and Mourão 2019), as well as its relatively easy, API–driven (application programming interface) access for researchers (Özkula et al. 2022). Especially during elections, Twitter has become a focal point for both campaigners and researchers (Conway et al. 2015; Jungherr 2016; Zhang et al. 2018). The EPINetz Project’s Twitter Politicians Dataset aims to facilitate political science research on the platform by providing a curated list of politically relevant accounts. While use of the platform among the general population is relatively sparse in most non–Anglo-Saxon countries, including Germany (Newman et al. 2021), scholars have argued that its popularity among political elites, its relatively low adoption K The EPINetz Twitter Politicians Dataset 2021. A New Resource for the Study of the German... 531 cost, and few topical and formal restrictions to content make it especially wellsuited to study political communication. Not only has it been noted that Twitter communications of politicians can be a rich data source comparable to press statements (Sältzer 2022),but it has also been argued that the self-selection of politically interested ordinary users on Twitter comprises an especially valuable subset of the average population when it comes to tracking political debates (Barberá 2015). Furthermore, as with all social media data, it can be argued that its unobtrusive observation of communication yields a higher validity and circumvents certain forms of bias, such as social acceptance or party guidelines, that are present in surveys or (press) interviews (Barberá et al. 2015;Ceron2017). In this regard, research concerning the generational gap in social media and online news use is proving vital in assessing the potential for generalising these findings (Mangold et al. 2021). In political science, Twitter data have been employed to estimate the ideological positions of ordinary citizens (Barberá et al. 2015;Barberá2015), between parties (Ecker 2017) and in intraparty politics (Ceron 2017; Sältzer 2022). Furthermore, Twitter data have been utilised as a valid proxy for political issue attention among both citizens and political elites, enabling the study of issue ownership and politicians’ responsiveness to the media and public (Barberá et al. 2019; Ceron et al. 2020; Franzmann et al. 2020). The platform has also been under study for assessing different communication styles and campaign negativity between parties (Evans et al. 2014; Hegelich and Shahrezaye 2015; Russell 2018; Petkevic and Nai 2021). Connected to these questions is research concerned with the fragmentation of political communication into so-called echo chambers and their detrimental effects on the political climate (Barberá et al. 2015; Hegelich and Shahrezaye 2015; Vaccari et al. 2016). Likewise, the assessment of populist rhetoric on Twitter has been under special scrutiny (Ernst et al. 2017; Gründl 2020; Maurer and Diehl 2020), since social media is suspected to empower populist parties (Engesser et al. 2017; Jungherr et al. 2019b). Due to the high volume of political communication, the assumed representativeness of party positions, and their role as a democratic focal point, election campaigns are the focus in the majority of the aforementioned studies (Evans et al. 2014; Hegelich and Shahrezaye 2015; Vaccari et al. 2016;Ceron2017; Russell 2018; Ceron et al. 2020; Franzmann et al. 2020; Petkevic and Nai 2021). For the German context, a number of studies specifically analyse politicians’ use of the platform during federal election campaigns (Jürgens and Jungherr 2015;Schmidt2017,2021; Stier et al. 2018b). In this regard, questions of a changing landscape of political communication arise—both in general terms (Van Aelst et al. 2017; Borucki and Jun 2018)and with a focus on German political institutions (Murphy 2019). During the COVID-19 pandemic, these questions of governmental communication on Twitter have gained new urgency (Rufai and Bunce 2020; Rivas-De-roca et al. 2021). For all these research approaches, and despite the relatively liberal data access that Twitter grants to researchers, a major obstacle remains. The open architecture, low access barriers, and high anonymity of the platform pose a challenge when attempting to tie observations back to the institutionalised political system. While topic-driven sampling approaches through hashtags and other methods merit credit for producing valuable insights on how discourse on the platform works, critics may claim they have little resemblance to the actual political discourse due to, in K 532 T. König et al. many countries, the platform ’ s low use among the average population (Newman et al. 2021). At the same time, identifying relevant speakers, such as politicians, can prove challenging, as no central databases tracking their social media presences exist. Researchers interested in the communication of political institutions and politicians, then, would have to handpick the relevant Twitter accounts—a task too resourceintensive for many research projects. While some databases exist that compile extensive information on politicians and parliamentarians (Gerring et al. 2019; Göbel and Munzert 2021), they do not provide a comprehensive overview of their social media presence. There are, however, a few databases focussing on politicians’ presence on the Twitter platform in particular. Most notably, the Politicians on Social Media database1holds information on the Twitter accounts of more than 12,000 politicians from more than 154 countries (Haman and Školník 2021). With a more narrow, European focus, the Twitter Parliamentarian Database holds data for 27 countries and 6437 politicians (van Vliet et al. 2020). For the case of Germany, the GESIS institute released comprehensive datasets of candidates running in the 2013, 2017, and 2021 federal elections (Kaczmirek and Mayr 2015; Stier et al. 2018a; Sältzer et al. 2021). However, all these databases only hold information for parliamentarians elected (or running) on a national or federal level. We argue, however, that this only provides a small window into the communications of institutionalised politics on Twitter. For multilevel polities such as Germany, one has to assume that relevant debates are not limited to the federal level but take place on a state level as well. Moreover, depending on the issue at hand, they are also more or less intertwined with debates on the European level. However, a comprehensive database of relevant actors on a state, federal, and European Union (EU) level is still lacking for Germany. Furthermore, as with any database, without proper and constant maintenance, the data provided become outdated sooner rather than later, establishing a constant demand for up-to-date and well-maintained datasets—especially in the fast-moving world of digital communications. To cater to these needs, we present the EPINetz Twitter Politicians Dataset. As part of the EPINetz Project,2we curate and store information on the Twitter presence of all German parliamentarians on a federal, state, and EU level, as well as German political institutions such as ministries and parties. This up-to-date, hand-curated dataset provides information on a total of 2449 Twitter accounts, along with their office, party affiliation, and, where applicable, additional information such as their term of service, gender, and year of birth. The data provided were collected over the course of 2021, meaning that all parliamentarians and institutions in office during that year were collected. For parliaments reelected during this time, such as the federal parliament in September 2021, we provide data for both the pre-2021 and post- 2021 legislative period. While we do not provide actual tweet data, any researcher with access to Twitter data (such as the official API) can use this dataset to retrieve reliable data on German politicians’ Twitter activity during their terms of service. A fine-grained data structure allows for numerous analyses considering differences between the state and federal levels, certain regions, institutions, offices, and parties, 1https://www.politiciansonsocialmedia.com 2https://epinetz.de/ K The EPINetz Twitter Politicians Dataset 2021. A New Resource for the Study of the German... 533 as well as personal attributes such as age and gender. Finally, the comprehensive overview of accounts of federal politicians and institutions provides an up-to-date basis for comparative research on a national level. As such, our dataset will prove a valuable resource for research on, but not limited to, election campaigns, fragmentation, populism, ideological positioning, governmental and party communication, issue attention, and intermedia effects on and through Twitter, as well as questions regarding the influence of sociodemographic variables. In the following, we provide a detailed description of the EPINetz Twitter Politicians Dataset, its method of acquisition, and specific variables. Furthermore, we showcase several possible uses by providing a brief analysis of the 2021 German Federal Elections utilising our dataset. Finally, the conclusion provides further applications and avenues for research, as well as highlights the limitations of the data provided. The dataset can be accessed via the GESIS SowiDataNet|datorium (König et al. 2022).3 2 The EPINetz Twitter Politicians Dataset For the initial data acquisition, we built on a number of Twitter lists of German politicians in parliament that were curated and made publicly available by Martin Fuchs, aka “Wahlbeobachter” (election observer).4In order to assess this data, we relied on an early, nonpublic version of the Social Media Observatory’s Twitter Parliamentarian Database (Münch et al. 2021).5After the initial evaluation in early 2021, for every state or federal election thereafter, we scanned the official parliament databases for newly elected members and searched for these members’ Twitter accounts by hand. At the same time, we updated their terms of offices as needed, e.g., when voted out of parliament. Note that, due to the initial evaluation taking place in early 2021, office terms are tracked only for those legislative periods stretching into and starting in 2021 (see Online Appendix A for a comprehensive overview). Finally, in late 2021, after federal elections and numerous state elections had been held, we reviewed the data collected by comparing it with the now updated Twitter lists of German politicians in parliament curated by Martin Fuchs. This list of parliamentarians was then supplemented by an in-depth investigation into the Twitter accounts of German ministers, state secretaries, ministries, and parties on both federal and state levels. As to the EU level, given the multinational character of its institutions, we collected data for German Members of the European Parliament (MEPs) only. Manual data collection to retrieve accounts or resolve inconsistencies between data sources followed a standardised procedure: 1) retrieval of information on government or parliament members through official sources; 2) lookup of matching accounts through the Twitter website’s search function; and 3) evaluation of the account’s description, timeline, and displayed name to determine whether it represented the politician or institution in question. Whenever possible, we pri- 3https://doi.org/10.7802/2415 4http://www.hamburger-wahlbeobachter.de /https://twitter.com/wahl_beobachter 5Special thanks to Felix-Victor Münch and the team at the Social Media Observatory for granting us access to this work-in-progress database. K 534 T. König et al. Table 1 Variables in the EPINetz Twitter Politicians Dataset Variable Description ID A unique identifier, generated over user_id, region, institution, office, and party membership Official_name The official name of a person or institution as referred to in official documents or the press. Middle names and titles are dropped From Date of a person’s assumption of office. Only goes back to the last legislative period. Not available for parties and ministries Until Date of a person’s dismissal of office. Not available for ongoing offices. Never available for parties and ministries Party Party affiliation of a person or official party account. Not available for ministries Region The political region a person’s, party’s, or ministry’s activities are focused on. One of the German federal states, the federal level (Bund), or the European Union Institution The institutional affiliation of an account. State, federal, or European parliament for parliamentarians; state or federal government for minsters, ministries, and state secretaries; “Party” for official party accounts Office Political office associated with an account. Refers to their role as parliamentarian, minister, state secretary, or ministry. Accounts associated with parties are split into “Parliamentary Party Group” (for accounts representing parliamentary factions) and “Speaker” (for accounts representing party speakers) Twitter_name An account’s name as displayed on Twitter. Subject to change by users Twitter_handle An account’s Twitter handle (@handle). Subject to change by users User_ID An account’s unique Twitter ID. Cannot be changed by users Year_of_birth A person’s year of birth, if available. Never available for parties and ministries Gender A person’s self-ascribed gender, if available. Never available for parties and ministries Abgeordnetenwatch_ID A person’s ID on the platform Abgeordnetenwatch (Parliamentwatch), if available. Available only for parliamentarians and former parliamentarians Wikidata_ID The ID of the associated Wikidata page, if available oritised politicians’ official over personal accounts. Accounts that were inactive or superseded by another account remained in the database for archival reasons, e.g., when research interest requires retrieving tweets for a longer period of time.6In order to obtain additional sociodemographic variables, we used the Wikidata API.7 Specifically, we retrieved information on gender and year of birth for all individual persons, and the Abgeordnetenwatch (Parliamentwatch) ID for parliamentarians. When these variables were unavailable on Wikidata, we manually extracted this information from the Abgeordnetenwatch website instead.8 The EPINetz Twitter Politicians Dataset comes with a number of variables, allowing for a more fine-grained analysis. Table 1gives a comprehensive overview. 6This produces multiple account entries for politicians with multiple accounts, e.g., for the new Foreign Minister, Annalena Baerbock, whose old Twitter account was superseded and renamed as “@ABaerbockArchiv” when she assumed office. 7https://www.wikidata.org/. Note that we also use the API to supplement the existing entries with the social media information here collected, allowing researchers to access up-to-date information on German politicians’ Twitter accounts directly via Wikidata. 8https://www.abgeordnetenwatch.de/ K The EPINetz Twitter Politicians Dataset 2021. A New Resource for the Study of the German... 535 The ID makes sure the same Twitter account associated with different offices (e.g., when a politician is both a member of parliament and a minister), regions (e.g., changing or multiple parliamentary memberships), institutions (e.g., a change from the state to the federal government), or parties (e.g., changes in party membership) can be uniquely identified in its different roles. Due to this representation of multiple roles and contexts, filtering is advised when collecting data to avoid duplicating tweets. When collecting data, it is also strongly advised to utilise the unchanging User_ID rather than a user’s Twitter_handle. The date of reference for a change in offices (the From and Until variables) is the first session of the newly elected parliament after election or the formation of the new government (see Online Appendix A). Note that, while changes in governments (e.g., changes in ministerial positions during legislatures) are represented in our data, we do not comprehensively track changes in parliament or party affiliation in between these dates of reference. The Abgeordnetenwatch_ID represents a person’s unique identifier on the independent, nonpartisan platform Abgeordnetenwatch (Parliamentwatch). The website tracks incumbencies, candidacies, and voting records for all German parliamentarians. It also allows citizens to publicly pose questions to their representatives. The identifier was included to enable further research avenues, e.g., when trying to associate Twitter activity with voting behaviour or responsiveness to citizens’ questions. We also included the matching Wikidata ID (if available) for all accounts. Additional information on the distribution of variables in the data can be found in Online Appendix B. 3 Exemplary Analysis: German Federal Elections 2021 In this section, we present an exemplary study on the Twitter communication of German politicians during the federal election campaign 2021. Therefore, we collected the timelines (all tweets, including retweets, replies to other tweets, and their associated metadata) of all accounts in our dataset for the period between 19 April and 26 September. We thus let our research period start with the day of the nomination of Annalena Baerbock of the Green Party, the second nomination of a lead candidate after the Social Democratic Party (SPD) had already announced that Olaf Scholz was running for Chancellor. The research period “naturally” ends with the closing of ballot boxes at election day, 6p.m. We further restricted our sample to accounts related to the seven largest parties represented in the 19th German Bundestag and thus to the parties with a realistic outlook to win mandates in the 20th election. These were the Alternative for Germany (AfD; an extreme right-wing party), the Green Party (Bündnis 90/Die Grünen), the Christian Democratic Party (CDU), the Christian Social Union in Bavaria (CSU; the CDU’s Bavarian sister party), the Liberals (FDP), the Left Party (DIE LINKE), and the SPD. Accounts with no party affiliation, i.e., ministries, were dropped for this analysis. To collect the data, we K 536 T. König et al. Fig. 1 Twitter accounts (total parliamentarians in the dataset) compared to mandates (total of elected parliamentarians across all regions) by party affiliation used Twitter’s academic V2 API access, allowing for full archive searching.9Our data contain a total of 426,614 tweets for the observation period, 188,532 (44%) of which are retweets. 3.1 Twitter Usage by Parties During the Election Campaign Previous research has richly documented the activity differences between parties with respect to certain social media (Jürgens and Jungherr 2015;Schmidt2017, 2021). These can also be studied using our dataset. First, however, it is important to note that the distribution of Twitter accounts over parties is inherently unbalanced. Figure 1shows that there are significant differences between a party’s mandates in parliament and its delegates’ Twitter adoption. Politicians and parties vary in the use of Twitter, with lower shares of Twitter account holders and less regular usage among conservative parties. From this perspective, the CDU is especially underrepresented in our dataset. When measured by mandates across all 18 parliaments (regional, federal, EU), the CDU is clearly the largest political party in the country before the 2021 federal elections. However, with less than 50% of its elected politicians on Twitter and their comparatively low activity (see below), the party has relatively low visibility on the platform. This explainable imbalance in the sample should be kept in mind when interpreting the following findings. It shows that Twitter data are not 9See https://developer.twitter.com/en/products/twitter-api/academic-research for details on Twitter’s academic research access. We used the academictwitteR package for the R programming language to access thedata(BarrieandHo2021). K The EPINetz Twitter Politicians Dataset 2021. A New Resource for the Study of the German... 543 for the AfD confirms its classification as populist, while Die Linke’s slightly aboveaverage values underline its status as a borderline case. Generally speaking, populist messages are not extraordinarily prevalent among politicians on Twitter during the election campaign. These results confirm Gründl’s (2020) findings for the time period of 2014–2020.10 For an analysis of populist dynamics in the election campaign, Fig. 4shows the percentage of sentences with populist messages over time both as 30-day rolling means per party and as a fitted total average. We can see how, interestingly, the AfD spread more populist messages in June, with a downwards trend towards election day. This, and the observable spikes, suggests that shortterm issues determine the populist communication of this party more strongly than campaign dynamics. In contrast, for all other parties and the fitted average, we can observe varying increases of populist messages towards election day, reflecting an increasingly heated election battle with uncertain outcomes. This is especially true for the left party Die Linke which, fearing electoral losses, strongly increased its populist messages over the last 6 weeks of the election campaign. These results correspond with findings that propose that campaign negativity may correlate with an increasingly desperate campaign (Petkevic and Nai 2021), suggesting similar patterns for both phenomena. 4Conclusion With this research note, we gave an overview of the EPINetz Twitter Politicians Dataset, the variables contained, and the accounts tracked. To illustrate potential use cases for the dataset in political communication research, we presented an analysis of the 2021 German federal elections, providing exploratory insights into the Twitter use of politicians during the election period. We would like to encourage researchers to use the dataset and conduct research beyond the context of election campaigns and the various methods showcased here. Potential use cases are longitudinal studies on politicians’ tweet behaviour, additional uses for the sociodemographic variables provided, and utilisation of the linkages to additional data sources, such as Abgeordnetenwatch for the tracking of voting behaviour in parliament or Wikidata for the collection and analysis of additional data. Current limitations of the dataset are that it does not contain comprehensive tracking of parliamentarians’ mandates during legislative periods and that we provide data only for the legislative periods of 2021. We intend, however, to annually update the EPINetz Twitter Politicians Dataset to continually provide an up-to-date resource for researchers. In this regard, we are 10 When compared to Gründl’s (2020) results, lower populism in our sample can be explained by the extension of the data to all party politicians, rather than only official party accounts and leaders. Online Appendix C shows the number of populist messages for a subset of accounts mirroring Gründl’s sample. As expected, these numbers show generally higher amounts of populism during the election campaign with its more heated political debates. The CDU’s higher number of populist messages during the election campaign for all subsets can be explained by the Merkel era of relative stability and calm observed by Gründl (2014–2020), being superseded by Laschet’s increasingly desperate (and ultimately failed) 2021 election campaign. K 544 T. König et al. looking forward to seeing exciting new applications of the dataset both within and outside the field of political science. Supplementary Information The online version of this article (https://doi.org/10.1007/s11615-022- 00405-7) contains supplementary material, which is available to authorized users. Acknowledgements We would like to thank our research assistants Nele Hartwig, Quentin Bukold, and Marco Gronewald for their valuable contributions in selecting and evaluating the accounts collected in the dataset. Funding This research was funded by the Klaus Tschira Foundation as part of the EPINetz project. Funding Open Access funding enabled and organized by Projekt DEAL. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4. 0/. Conflict of interest T. König, W.J. Schünemann, A. Brand, J. Freyberg, and M. Gertz declare that they have no competing interests. References Barberá, Pablo. 2015. Birds of the same feather tweet together: Bayesian ideal point estimation using twitter data. Political Analysis 23:76–91. https://doi.org/10.1093/pan/mpu011. Barberá, Pablo, John T. Jost, Jonathan Nagler, Joshua A. Tucker, and Richard Bonneau. 2015. Tweeting from left to right: is online political communication more than an echo chamber? Psychological Science 26:1531–1542. https://doi.org/10.1177/0956797615594620. Barberá, Pablo, Andreu Casas, Jonathan Nagler, Patrick J. Egan, Richard Bonneau, John T. Jost, and Joshua A. Tucker. 2019. Who leads? Who follows? Measuring issue attention and agenda setting by legislators and the mass public using social media data. American Political Science Review 113:883–901. https://doi.org/10.1017/S0003055419000352. Barrie, Christopher, and Justin Chun-ting Ho. 2021. academictwitteR: an R package to access the twitter academic research product track v2 API endpoint. Journal of Open Source Software 6:3272. https:// doi.org/10.21105/joss.03272. Borucki, Isabelle, and Uwe Jun. 2018. Regierungskommunikation im Wandel – Politikwissenschaftliche Perspektiven. In Regierungskommunikation und staatliche Öffentlichkeitsarbeit: Implikationen des technologisch induzierten Medienwandels, ed. Juliana Raupp, Jan Niklas Kocks, and Kim Murphy, 25–46. Wiesbaden: Springer. https://doi.org/10.1007/978-3-658-20589-8_3. Bossetta, Michael. 2018. The digital architectures of social media: comparing political campaigning on Facebook, Twitter, Instagram, and Snapchat in the 2016U.S. election. Journalism & Mass Communication Quarterly 95:471–496. https://doi.org/10.1177/1077699018763307. Ceron, Andrea. 2017. Intra-party politics in 140 characters. Party Politics 23:7–17. https://doi.org/10.1177/ 1354068816654325. Ceron, Andrea, Luigi Curini, and Wiebke Drews. 2020. Short-term issue emphasis on twitter during the 2017 German election: a comparison of the economic left-right and socio-cultural dimensions. German Politics https://doi.org/10.1080/09644008.2020.1836161. Conway, Bethany A., Kate Kenski, and Di Wang. 2015. The rise of twitter in the political campaign: searching for Intermedia agenda-setting effects in the presidential primary. Journal of Computer- Mediated Communication 20:363–380. https://doi.org/10.1111/jcc4.12124. K The EPINetz Twitter Politicians Dataset 2021. A New Resource for the Study of the German... 545 Dagoula, Chrysi. 2019. Mapping political discussions on twitter: where the elites remain elites. Media and Communication 7:225–234. https://doi.org/10.17645/mac.v7i1.1764. Dimitrova, Daniela V., and Jörg Matthes. 2018. Social media in political campaigning around the world: theoretical and methodological challenges. Journalism & Mass Communication Quarterly 95:333–342. https://doi.org/10.1177/1077699018770437. Ecker, Alejandro. 2017. Estimating policy positions using social network data: cross-validating position estimates of political parties and individual legislators in the Polish parliament. Social Science Computer Review 35:53–67. https://doi.org/10.1177/0894439315602662. Engesser, Sven, Nayla Fawzi, and Anders Olof Larsson. 2017. Populist online communication: introduction to the special issue. Information, Communication & Society 20:1279–1292. https://doi.org/10. 1080/1369118X.2017.1328525. Ernst, Nicole, Sven Engesser, Florin Büchel, Sina Blassnig, and Frank Esser. 2017. Extreme parties and populism: an analysis of Facebook and Twitter across six countries. Information, Communication & Society 20:1347–1364. https://doi.org/10.1080/1369118X.2017.1329333. Evans, Heather K., Victoria Cordova, and Savannah Sipole. 2014. Twitter style: an analysis of how house candidates used twitter in their 2012 campaigns. PS: Political Science & Politics 47:454–462. https:// doi.org/10.1017/S1049096514000389. Franzmann, Simon T., Heiko Giebler, and Thomas Poguntke. 2020. It’s no longer the economy, stupid! Issue yield at the 2017 German federal election. West European Politics 43:610–638. https://doi.org/ 10.1080/01402382.2019.1655963. Gerring, John, Erzen Oncel, Kevin Morrison, and Daniel Pemstein. 2019. Who rules the world? A portrait of the global leadership class. Perspectives on Politics 17:1079–1097. https://doi.org/10.1017/ S1537592719000744. Göbel, Sascha, and Simon Munzert. 2021. The comparative legislators database. British Journal of Political Science https://doi.org/10.1017/S0007123420000897. Gründl, Johann. 2020. Populist ideas on social media: a dictionary-based measurement of populist communication. New Media & Society https://doi.org/10.1177/1461444820976970. Haman, Michael, and Milan Školník. 2021. Politicians on social media. The online database of members of national parliaments on twitter. Profesional de la Información https://doi.org/10.3145/epi.2021.mar. 17. Heft, Annett, Curd Knüpfer, Susanne Reinhardt, and Eva Mayerhöffer. 2021. Toward a transnational information ecology on the right? Hyperlink networking among right-wing digital news sites in europe and the United States. The International Journal of Press/Politics 26:484–504. https://doi.org/10. 1177/1940161220963670. Hegelich, Simon, and Morteza Shahrezaye. 2015. The communication behavior of German MPs on twitter: preaching to the converted and attacking opponents. European Policy Analysis 1:155–174. https://doi. org/10.18278/epa.1.2.8. Jungherr, Andreas. 2016. Twitter use in election campaigns: a systematic literature review. Journal of Information Technology & Politics 13:72–91. https://doi.org/10.1080/19331681.2015.1132401. Jungherr, Andreas, Oliver Posegga, and Jisun An. 2019a. Discursive power in contemporary media systems: a comparative framework. The International Journal of Press/Politics 24:404–425. https://doi. org/10.1177/1940161219841543. Jungherr, Andreas, Ralph Schroeder, and Sebastian Stier. 2019b. Digital media and the surge of political outsiders: explaining the success of political challengers in the United States, Germany, and China. Social Media + Society https://doi.org/10.1177/2056305119875439. Jungherr, Andreas, Gonzalo Rivero, and Daniel Gayo-Avello. 2020. Retooling politics: how digital media are shaping democracy. Cambridge: Cambridge University Press. https://doi.org/10.1017/ 9781108297820. Jürgens, Pascal, and Andreas Jungherr. 2015. The use of twitter during the 2009 German national election. German Politics 24:469–490. https://doi.org/10.1080/09644008.2015.1116522. Kaczmirek, Lars, and Philipp Mayr. 2015. German Bundestag elections 2013: twitter usage by electoral candidates. Cologne: GESIS Data Archive. https://doi.org/10.4232/1.12319. Knüpfer, Curd, Matthias Hoffmann, and Vadim Voskresenskii. 2020. Hijacking MeToo: transnational dynamics and networked frame contestation on the far right in the case of the ‘120 decibels’ campaign. Information, Communication & Society https://doi.org/10.1080/1369118X.2020.1822904. König, Tim, Alexander Brand, Julian Freyberg, Wolf J. Schünemann, and Michael Gertz. 2022. EPINetz Twitter Politicians 2021.https://doi.org/10.7802/2415. SowiDataNet|datorium. K 546 T. König et al. Langer, Ana Ines, and Johannes B. Gruber. 2021. Political agenda setting in the hybrid media system: why legacy media still matter a great deal. The International Journal of Press/Politics 26:313–340. https:// doi.org/10.1177/1940161220925023. Mangold, Frank, Sebastian Stier, Johannes Breuer, and Michael Scharkow. 2021. The overstated generational gap in online news use? A consolidated infrastructural perspective. New Media & Society https://doi.org/10.1177/1461444821989972. Margetts, Helen. 2019. Rethinking democracy with social media. The Political Quarterly 90:107–123. https://doi.org/10.1111/1467-923X.12574. Maurer, Peter, and Trevor Diehl. 2020. What kind of populism? Tone and targets in the Twitter discourse of French and American presidential candidates. European Journal of Communication 35:453–468. https://doi.org/10.1177/0267323120909288. Molyneux, Logan, and Rachel R. Mourão. 2019. Political journalists’ normalization of twitter. Journalism Studies 20:248–266. https://doi.org/10.1080/1461670X.2017.1370978. Münch, Felix Victor, Mani Levian Asli, Khandoker Tanjim Ahammad, Gregor Wiedemann, Jan Rau, Jan- Hinrik Schmidt, and Cornelius Puschmann. 2021. Social media observatory wiki.https://doi.org/10. 6084/m9.figshare.14465337.v3. Social Media Observatory (SMO). Murphy, Kim. 2019. Government communications in a digital age. A comparative study of online government communications in Germany and great britain. Baden-Baden: Nomos. https://doi.org/10.5771/ 9783845298030. Murthy, Dhiraj. 2012. Towards a sociological understanding of social media: theorizing twitter. Sociology 46:1059–1073. https://doi.org/10.1177/0038038511422553. Newman, Nic, Richard Flechter, Anne Schulz, Simge Andi, Craig T. Robertson, and Rasmus Kleis Nielsen. 2021. Reuters institute digital news report 2021. Oxford: Reuters Institute for the Study of Journalism. Özkula, Suay M., Paul J. Reilly, and Jenny Hayes. 2022. Easy data, same old platforms? A systematic review of digital activism methodologies. Information, Communication & Society https://doi.org/10. 1080/1369118X.2021.2013918. Persily, Nathaniel, and Joshua A. Tucker. 2020. Social media and democracy: the state of the field, prospects for reform. Cambridge: Cambridge University Press. https://doi.org/10.1017/9781108890 960. Petkevic, Vladislav, and Alessandro Nai. 2021. Political attacks in 280 characters or less: a new tool for the automated classification of campaign negativity on social media. American Politics Research https:// doi.org/10.1177/1532673X211055676. Rivas-De-roca, Rubén, Mar García-Gordillo, and José L. Rojas-Torrijos. 2021. Communication strategies on twitter and institutional websites in the covid-19 second wave: Analysis of the governments of Germany, Spain, Portugal and the United Kingdom. Revista Latina de Comunicacion Social 79:49–72. https://doi.org/10.4185/RLCS-2021-1517. Rufai, Sohaib R., and Catey Bunce. 2020. World leaders’ usage of Twitter in response to the COVID-19 pandemic: a content analysis. Journal of Public Health 42:510–516. https://doi.org/10.1093/pubmed/ fdaa049. Russell, Annelise. 2018. U.S. Senators on twitter: asymmetric party rhetoric in 140 characters. American Politics Research 46:695–723. https://doi.org/10.1177/1532673X17715619. Sältzer, Marius. 2022. Finding the bird’s wings: dimensions of factional conflict on twitter. Party Politics 28:61–70. https://doi.org/10.1177/1354068820957960. Sältzer, Marius, Sebastian Stier, Joscha Bäuerle, Manuela Blumenberg, Valeriya Mechkova, Daniel Pemstein, Brigitte Seim, and Steven Wilson. 2021. Twitter accounts of candidates in the German federal election 2021. Cologne: GESIS Data Archive. https://doi.org/10.4232/1.13789. Schmidt, Jan-Hinrik. 2017. Facebook- und Twitter-Nutzung der Kandidierenden zur Bundestagswahl 2017. Verbreitung, Aktivität und Informationsquellen. Media Perspektiven 2017:616–629. Schmidt, Jan-Hinrik. 2021. Facebook- und Twitter-Nutzung der Kandidierenden zur Bundestagswahl 2021. Verbreitung, Aktivität und Informationsquellen. Media Perspektiven 2021:639–653. Stier, Sebastian, Arnim Bleier, Malte Bonart, Fabian Mörsheim, Mahdi Bohlouli, Margarita Nizhegorodov, Lisa Posch, Jürgen Maier, Tobias Rothmund, and Steffen Staab. 2018a. Social media monitoring for the German federal election 2017. Cologne: GESIS Data Archive. https://doi.org/10.4232/1.12992. Stier, Sebastian, Arnim Bleier, Haiko Lietz, and Markus Strohmaier. 2018b. Election campaigning on social media: politicians, audiences, and the mediation of political communication on Facebook and twitter. Political Communication 35:50–74. https://doi.org/10.1080/10584609.2017.1334728. Stier, Sebastian, Caterina Froio, and Wolf J. Schünemann. 2021. Going transnational? Candidates’ transnational linkages on twitter during the 2019 European parliament elections. West European Politics 44:1455–1481. https://doi.org/10.1080/01402382.2020.1812267. K The EPINetz Twitter Politicians Dataset 2021. A New Resource for the Study of the German... 547 Su, Yan, and Porismita Borah. 2019. Who is the agenda setter? Examining the intermedia agenda-setting effect between Twitter and newspapers. Journal of Information Technology & Politics 16:236–249. https://doi.org/10.1080/19331681.2019.1641451. Vaccari, Cristian, Augusto Valeriani, Pablo Barberá, John T. Jost, Jonathan Nagler, and Joshua A. Tucker. 2016. Of echo chambers and Contrarian clubs: exposure to political disagreement among German and Italian users of twitter. Social Media + Society https://doi.org/10.1177/2056305116664221. Van Aelst, Peter, Jesper Strömbäck, Toril Aalberg, Frank Esser, Claes de Vreese, Jörg Matthes, David Hopmann, Susana Salgado, Nicolas Hubé, Agnieszka St˛epi´ nska, Stylianos Papathanassopoulos, Rosa Berganza, Guido Legnante, Carsten Reinemann, Tamir Sheafer, and James Stanyer. 2017. Political communication in a high-choice media environment: a challenge for democracy? Annals of the International Communication Association 41:3–27. https://doi.org/10.1080/23808985.2017.1288551. Van Vliet, Livia, Petter Törnberg, and Justus Uitermark. 2020. The twitter parliamentarian database: analyzing twitter politics across 26 countries. PLoS ONE https://doi.org/10.1371/journal.pone.0237073. Zhang, Yini, Chris Wells, Song Wang, and Karl Rohe. 2018. Attention and amplification in the hybrid media system: the composition and activity of Donald Trump’s Twitter following during the 2016 presidential election. New Media & Society 20:3161–3182. https://doi.org/10.1177/1461444817744390. K