Intra-Party Preference Heterogeneity and Faction Membership in the 15th German Bundestag: A Computational Text Analysis of Parliamentary Speeches

Abstract

In a broad range of research in comparative politics, political parties are conceptualised as unitary actors with consistent preferences. We depart from this sometimes accurate, at other times overly strong assumption by studying patterns of intra-party heterogeneity of preferences within parliamentary parties in the German Bundestag from 2002–05. For this purpose, we use the Wordscores method, a form of computational text analysis, to estimate policy positions of 453 individual legislators based on plenary speeches. We then study the link between intra-party faction membership and expressed policy positions. We find that there is a limited, but consistent effect of intra-party factionalism in the German Bundestag. According to random effects ANOVA, faction membership determines about 3 per cent of the variance of positions on economic policy in the present study.

Acknowledgements

We would like to thank the participants in the expert survey, and Simon Munzert for research assistance.

Notes

Patrick Köllner and Matthias Basedau, ‘Factionalism in Political Parties: An Analytical Framework for Comparative Studies’, Working Paper Global and Area Studies No. 12 (Hamburg: German Overseas Institute, 2005), pp.8f.

Michael Laver, Kenneth Benoit and John Garry, ‘Extracting Policy Positions from Political Texts Using Words as Data’, American Political Science Review 97/2 (2003), pp.311–31.

Patrick Köllner and Matthias Basedau, ‘Faktionalismus in politischen Parteien’, in Patrick Köllner, Matthias Basedau and Gero Erdmann (eds.), Innerparteiliche Machtgruppen. Faktionalismus im internationalen Vergleich (Frankfurt a.M.: Campus, 2006), p.9; for an earlier overview see Kenneth Janda, ‘Comparative Political Parties: Research and Theory’, in Ada W. Finifter (ed.), Political Science: The State of the Discipline II (Washington, DC: American Political Science Association, 1993), pp.163–91.

Giovanni Sartori, Parties and Party Systems: A Framework for Analysis (Cambridge: Cambridge University Press, 1976), pp.76–9.

Patrick Köllner, ‘Faktionen in der Liberaldemokratischen Partei Japans: Ursachen, Charakteristika und Konsequenzen’, in Köllner et al. (eds.), Innerparteiliche Machtgruppen, pp.247–76; Joachim Betz, ‘Faktionalismus in indischen Parteien’, in Köllner et al. (eds.), Innerparteiliche Machtgruppen, pp.275–302; Kim Eric Bettcher, ‘Factions of Interest in Japan and Italy. The Organizational and Motivational Dimension of Factionalism’, Party Politics 11/3 (2005), pp.339–58.

E.g. Ferdinand Müller-Rommel, Innerparteiliche Gruppierungen in der SPD. Eine empirische Studie über informell-organisierte Gruppierungen von 1969–1980 (Opladen: Westdeutscher Verlag, 1982).

Raphael Zariski, ‘Party Factions and Comparative Politics: Some Preliminary Observations’, Midwest Journal of Political Science 4/1 (1960), pp.26–51.

Köllner and Basedau, ‘Factionalism in Political Parties’.

Köllner et al. (eds.), Innerparteiliche Machtgruppen.

Gero Erdmann, ‘Zur Empirie innerparteilicher Machtgruppen im internationalen Vergleich’, in Köllner et al. (eds.), Innerparteiliche Machtgruppen, pp.333–60.

Gary W. Cox and Frances Rosenbluth, ‘Factional Competition for the Party Endorsement: The Case of Japan's Liberal Democratic Party’, British Journal of Political Science 26/2 (1996), pp.259–69; Gary W. Cox, Francis McCall Rosenbluth and Michael F. Thies, ‘Electoral Rules, Career Ambitions and Party Structure: Comparing Factions in Japan's Upper and Lower House’, American Journal of Political Science 44/1 (2000), pp.115–22; Scott Morgenstern, ‘Organized Factions and Disorganized Parties. Electoral Incentives in Uruguay’, Party Politics 7/2 (2001), pp.235–56.

Thomas König, ‘The Scope for Policy Change after the 2005 Election: Veto Players and Intra-Party Decision Making’, German Politics 15/4 (2006), pp.520–32; Marc Debus and Thomas Bräuninger, ‘Intra-Party Factions and Coalition Bargaining in Germany’, in Daniela Giannetti and Kenneth Benoit (eds.), Intra-Party Politics and Coalition Governments (London: Routledge, 2008), pp.121–45; Reimut Zohlnhöfer, ‘Destination Anywhere? The German Red–Green Government's Inconclusive Search for a Third Way in Economic Policy’, German Politics 13/1 (2004), pp.106–31.

George Tsebelis, Veto Players: How Political Institutions Work (Princeton, NJ: Princeton University Press, 2002).

Ibid., pp.38ff., p.63

König, ‘The Scope for Policy Change’, p.520.

Debus and Bräuninger, ‘Intra-Party Factions’.

Laver et al., ‘Extracting Policy Positions’.

Michael Laver and Kenneth Benoit, ‘Locating TDs in Policy Spaces: Wordscoring Dáil Speeches’, Irish Political Studies 17/1 (2002), pp.59–73.

Ibid.

Daniela Giannetti and Michael Laver, ‘Policy Positions and Jobs in the Government’, European Journal of Political Research 44/1 (2005), pp.91–120.

Giannetti and Laver (eds.), Intra-Party Politics, pp.146–68.

Wolfgang Ismayr, Der Deutsche Bundestag im politischen System der Bundesrepublik Deutschland (Opladen: Leske + Budrich, 2006), pp.110f.

Ibid., p.108; Kathrin Dümig, Matthias Trefs and Reimut Zohlnhöfer, ‘Die Faktionen der CDU: Bändigung durch institutionelle Einbindung?’, in Köllner et al. (eds.), Innerparteiliche Machtgruppen, pp.99–130.

Thomas Poguntke, ‘Parties in a Legalistic Culture: The Case of Germany’, in Richard S. Katz and Peter Mair (eds.), How Parties Organize. Change and Adaption in Party Organizations in Western Democracies (London: Sage, 1994), pp.201f.

Emil J. Kirchner and David Broughton, ‘The FDP in the Federal Republic of Germany: The Requirements of Survival and Success’, in Emil J. Kirchner (ed.), Liberal Parties in Western Europe (Cambridge: Cambridge University Press, 1988), pp.63f.

Saskia Richter, ‘Identitätsstifter: Die innerparteilichen Gruppen der deutschen Grünen’, in Köllner et al. (eds.), Innerparteiliche Machtgruppen, pp.131–56.

Kirchner and Broughton, ‘The FDP’, pp.63f.

The questionnaire is available from the corresponding author, Julian Bernauer. Email: [email protected]

These 547 are the MPs who delivered speeches related to either of five policy dimensions (including EU, foreign and domestic security policy) selected in the context of broader research.

The questionnaire on the SPD was actually filled in by a staff member.

Source is Kürschners Volkshandbuch, Deutscher Bundestag: 15. Wahlperiode (Darmstadt: NDV, 2003).

Laver et al., ‘Extracting Policy Positions’, p.311.

Ibid., pp.313–19.

Ibid., p.314.

Kenneth Benoit and Michael Laver, Party Policy in Modern Democracies (London: Routledge, 2006), p.125.

Franz U. Pappi and Susumu Shikano, ‘Ideologische Signale in den Wahlprogrammen der deutschen Bundestagsparteien 1980 bis 2002’, Arbeitspapiere Mannheimer Zentrum für europäische Sozialforschung 76, Mannheim (2004), p.2; Seymour M. Lipset and Stein Rokkan, ‘Cleavage Structures, Party Systems and Voter Alignments: An Introduction’, in Seymour M. Lipset and Stein Rokkan (eds.), Cleavage Structures, Party Systems and Voter Alignments: Cross National Perspectives (London: The Free Press, 1967), pp.1–64; Michael Laver and Ben W. Hunt, Policy and Party Competition (London: Routledge, 1992).

Benoit and Laver, Party Policy, pp.180f., 213.

Laver and Benoit, ‘Locating TDs’, p.64.

In the FDP parliamentary party, Guido Westerwelle's share of plenary speaking time clearly exceeded that of the formal chair Wolfgang Gerhardt. He also regularly spoke first in major debates, which indicates his de facto leadership in the FDP parliamentary party. Therefore, Westerwelle's speeches instead of Gerhardt's are used as reference texts. Bündnis 90/Die Grünen traditionally have two PPG leaders. Their speeches are combined, since each of them specialised in certain policy areas and spoke in the related debates. The PDS did not have a PPG status in the 15th Bundestag and consequently no PPG leaders.

Giannetti and Laver, ‘Policy Positions and Jobs’, p.103.

Benoit and Laver, Party Policy, pp.180f.

The records of the 139th plenary session could not be analysed due to poor quality of the data file. Difficulties arose concerning the removal of Zwischenfragen (questions to the speaker). These occur frequently during speeches and could only be partly removed. A systematic cleansing would be advantageous but is very time-consuming given the thousands of speeches. The extent of the ‘noise’ should not be overestimated, though. In a random sample of 15 speeches containing 23,151 words, only two Zwischenfragen with 157 words were found. Also not removed are comments in the plenary records, including the announcement of the speaker by the president or vice president, notes of reactions such as applause or laughter and the occurrence of heckling. These comments are likely to be equally distributed across all speeches and should therefore have limited influence on policy positions. Finally, floor interventions (Kurzinterventionen) during debates are not considered.

Ismayr, ‘Der Deutsche Bundestag’, pp.315f.

Not considered are question times (Fragestunden), which do not have the character of debates. The questions asked by MPs are rather short. Nevertheless, they could reveal interesting information on the policy concerns of MPs, since questioning the government is one of the few remaining rights granted to individual MPs. They are disregarded since their structure is different from debates.

Wordscores software can be downloaded from: http://www.tcd.ie/Political Science/wordscores/software.html. For the economic policy dimension, the 20,000 most frequent unique words were used in the calculation of positions on this dimension. This restriction was necessary due to insufficient computing power. Overall, about 100,000 unique words are found in all texts on economic, finance and welfare policy, many of them appearing very rarely. Only words which appear both in reference and virgin texts add certainty to the position estimates of virgin texts. But, since both position estimates and standard errors are weighted by word frequencies, dropping seldom used words has limited influence on these. These considerations are confirmed by recomputation of the scores for economic and welfare policy using the 10,000 most frequent unique words only, therefore dropping 10,000 even more frequently used unique words. While the mean number of total words in virgin texts is reduced by only 79 words to 9,530 words, the percentage of scored words rises on average from 94.4 to 96.5. Some party-unspecific movement of policy positions does occur, as the mean absolute shift of policy positions is 0.22 scale points. Standard errors change only incrementally, indicating that the quality of the estimates remains largely unchanged. While it is not desirable to ignore even very rarely spoken words, the strategy is unlikely to bias the results substantially.

Helmut Pruscha, Statistisches Methodenbuch. Verfahren, Fallstudien, Programmcodes (Berlin: Springer, 2006); for a political science application see Scott Morgenstern and Richard F. Potthoff, ‘The Components of Elections: District Heterogeneity, District-Time Effects and Volatility’, Electoral Studies 24/1 (2005), pp.17–40.

This also means that partial sums of squares do not ad up to the model sum of squares in the ANOVA.

Tom A.B. Snijders and Roel J. Bosker, Multilevel Analysis: An Introduction to Basic and Advanced Multilevel Modeling (London: Sage, 1999); Andrew Gelman and Jennifer Hill, Data Analysis Using Regression and Multilevel/Hierarchical Models (Cambridge: Cambridge University Press, 2007).

Snijders and Bosker, Multilevel Analysis.

Ibid., p.48.

Josef Schmid, Die CDU: Organisationsstrukturen, Politiken und Funktionsweisen einer Partei im Föderalismus (Opladen: Leske + Budrich, 1990).

Torbjörn Bergman, Wolfgang C. Müller, Kaare Str⊘m and Magnus Blomgren, ‘Democratic Delegation and Accountability: Cross-national Patterns’, in Kaare Str⊘m, Wolfgang C. Müller and Torbjörn Bergman (eds.), Delegation and Accountability in Parliamentary Democracies (Oxford: Oxford University Press, 2003), p.128.

Ismayr, ‘Der Deutsche Bundestag’, p.47.