The Nexus between National Party Preferences and State Elections – A Long-Term Perspective

Abstract

State elections are one of the most prominent features of Germany's multi-level political system. The prevailing view is that the standing of the national governing parties has a major influence on the results of state elections. Still, two perspectives on the nexus between national party preferences and state election results have so far received less scholarly attention: First, it remains unclear if short-term changes in the national government's popularity also have an effect on state election results. Second, do results of state elections also influence the standing of the national government? A reason for this might be that short-term factors are becoming more important for government evaluation and vote choices. This paper responds to these two questions by examining the nexus between state elections and the standing of the national government in a long-term perspective from 1977 to 2005 and by means of cross-sectional and time-series analyses.

ACKNOWLEDGEMENTS

I would like to thank Thorsten Faas and two anonymous referees for helpful comments. The remaining mistakes are solely mine.

ABOUT THE AUTHOR

Dr Evelyn Bytzek is Lecturer in the Department of Political Science at the University of Koblenz-Landau. She received her PhD in 2007 from the University of Mannheim. Her fields of research are electoral behaviour, political communication and public opinion.

Notes

1. K. Reif and H. Schmitt, ‘Nine Second-Order National Elections: A Conceptual Framework for the Analysis of European Election Results’, European Journal of Political Research 8 (1980), pp.3–44.

2. See K. Völkl, K.-U. Schnapp, E. Holtmann and O.W. Gabriel, ‘Zum Einfluss der Bundespolitik auf Landtagswahlen: theoretischer Rahmen und Analysemodelle’, in K. Völkl, K.-U. Schnapp, E. Holtmann and O.W. Gabriel (eds), Wähler und Landtagswahlen in der Bundesrepublik Deutschland (Baden-Baden: Nomos, 2008), pp.9–36 for an overview.

3. Explaining and examining the mutual links between the national government's popularity and state election results directs attention to the problem of endogeneity in these kinds of analyses. Still, since both phenomena are driven by many factors besides the respective other, the endogeneity problem is not too severe and can thus be ignored.

4. R. Dinkel, ‘Der Zusammenhang zwischen Bundes- und Landtagswahlergebnissen’, Politische Vierteljahresschrift 18 (1977), pp.348–59.

5. E.g. J.E. Mueller, ‘Presidential Popularity from Truman to Johnson’, American Political Science Review 64 (1970), pp.18–34; J.A. Stimson, ‘Public Support for American Presidents: A Cyclical Model’, Public Opinion Quarterly 40 (1976), pp.1–21.

6. Völkl et al., ‘Zum Einfluss der Bundespolitik auf Landtagswahlen’, p.31.

7. D. Hough and C. Jeffery, ‘Landtagswahlen: Bundestestwahlen oder Regionalwahlen?’, Zeitschrift für Parlamentsfragen 34 (2003), pp.79–94.

8. E. Bytzek, Ereignisse und ihre Wirkung auf die Popularität von Regierungen. Von der Schleyer-Entführung zur Elbeflut (Baden-Baden: Nomos, 2007); E. Bytzek, ‘Questioning the Obvious: Political Events and Public Opinion on the Government's Standing in Germany 1977–2003’, International Journal of Public Opinion Research 23 (2011), pp.406–36.

9. S. Iyengar and D.R. Kinder, News that Matter. Television and American Opinion (Chicago/London: Chicago University Press, 1987), pp.63–5.

10. S. Burkhart, ‘Parteipolitikverflechtung. Über den Einfluss der Bundespolitik auf Landtagswahlentscheidungen von 1976 bis 2000’, Politische Vierteljahresschrift 46 (2005), pp.14–38.

11. Reif and Schmitt, ‘Nine Second-Order National Elections’.

12. C.J. Anderson, Blaming the Government. Citizens and the Economy in Five European Democracies (Armonk, NY/London: Sharpe, 1995), p.95.

13. I. McAllister, ‘The Personalization of Politics’, in H.-D. Klingemann and R.J. Dalton (eds), Oxford Handbook of Political Behavior (Oxford: Oxford University Press, 2007), pp.571–84.

14. Compare to Anderson, Blaming the Government.

15. In a study by Anderson (ibid., p.102).

16. Supported by ibid.

17. R. Beckmann, P. Trein and S. Walter, ‘Dominanz der Ökonomie: Entscheidet die Wirtschaftslage Wahlen?’, in E. Bytzek and S. Roßteutscher (eds), Der unbekannte Wähler? Mythen und Fakten über das Wahlverhalten der Deutschen (Frankfurt am Main: Campus, 2011), pp.231–52; C. Kellermann and H. Rattinger, ‘Wirtschaftslage, Arbeitslosigkeit und zugeschriebene Regierungsverantwortung als Bestimmungsfaktoren der Wahlverhaltens bei den Bundestagswahlen 2002 und 2005’, in F. Brettschneider, O. Niedermayer and B. Weßels (eds), Die Bundestagswahl 2005. Analysen des Wahlkampfes und der Wahlergebnisse (Wiesbaden: VS Verlag für Sozialwissenschaften, 2007), pp.369–93.

18. ZA2391, the data set is available from http://zacat.gesis.org/webview/index.jsp?object=http://134.95.45.58:80/obj/fCatalog/Catalog13

19. Bundeswahlleiter, Wahl zum 17. Deutschen Bundestag am 27. September 2009. Heft 1: Ergebnisse und Vergleichszahlen früherer Bundestags-, Europa- und Landtagswahlen sowie Strukturdaten für die Bundestagswahlkreise (Wiesbaden: SFG Servicecenter Fachverlage, 2009).

20. P.D. Allison, Missing Data (Thousand Oaks, CA et al.: Sage, 2002); W.F. Velicer and S.M. Colby, ‘A Comparison of Missing-Data Procedures for ARIMA Time-Series Analysis’, Educational and Psychological Measurement 65 (2005), pp.596–615.

21. See M.S. Lewis-Beck and M. Stegmaier, ‘Economic Determinants of Electoral Outcomes’, Annual Review of Political Science 3 (2002), pp.183–219, for an overview.

22. Bytzek, Ereignisse, p.146.

23. Another solution for this problem is error correction models which help to disentangle long- and short-term trends between dependent and independent variables. Still, the focus of this analysis does not lie on the relation between the national government's popularity and unemployment, but on the effect on state election results, which are short-term in nature. Therefore simpler ARIMA models are used.

24. Technically, the unemployment series was adjusted to a one-lag autoregressive term, which makes the resulting series stationary.

25. Computing several models to test the influence of the unemployment rate on the national government's popularity with different lag structures shows that a three-month lag performs best and thus is part of the model.

26. Technically, this is a lagged dependent variable with lag 1.

27. J. Peter, ‘Medien-Priming – Grundlagen, Befunde und Forschungstendenzen’, Publizistik 47 (2002), pp.21–44.