The perennial success of the German Greens

Abstract

The German Greens achieved a record result in the federal elections of 2009. Using both qualitative and quantitative methods, it is argued that this excellent result was not an isolated phenomenon but is in line with a long-term growth of Green electoral support that has a strong generational basis. A ‘feminisation’ and ‘greying’ of Green voters is also apparent. Despite the party's effort to emphasise economic and social issues in its campaigning, the chief factors explaining Green voting remain environmental concern and opposition to nuclear energy.

Keywords:

• Green Party
• Germany
• voting
• political generations

Notes

 1. For detailed documentation of all election results and polling figures, see the website http://www.wahlrecht.de [Accessed 7 July 2011].

 2. For a detailed explanation of the German electoral system, see the website http://www.wahlrecht.de/english/bundestag.htm [Accessed 7 November 2010].

 3. The events that are reported here have been widely covered in the German media; the discussion draws on the online editions of all major German newspapers – the Berlin daily Die Tageszeitung (TAZ) continues to be of particular value for the coverage of Green politics.

 4. While the results of the ‘official’ election statistics are based on a random sample of voters, the number of cases is so high that any sampling error is very small. The results produced can thus be taken as reflecting a very close approximation of voting choices according to sex and age.

 5. The data from the German post-election survey 2009 (GLES 1102, pre-release 1.2) were made available by GESIS – Leibniz-Institut für Sozialwissenschaften. It was collected by Hans Rattinger (University of Mannheim), Sigrid Roßteutscher (University of Frankfurt), Rüdiger Schmitt-Beck (University of Mannheim), and Bernhard Weßels (Science Centre Berlin) as part of the German Longitudinal Election Study in close cooperation with the German Society for Electoral Research (DGfW) and GESIS – Leibniz-Institut für Sozialwissenschaften. The data were documented and prepared by GESIS for secondary analysis, the research was funded by the German Research Foundation (DFG). The original collectors of the data and the relevant funding agencies and institutions bear no responsibility for uses of this dataset or for interpretations or inferences based upon such uses. For further information, cf. the DGfW website, http://www.dgfw.info/gles.php [Accessed 8 November 2010].

 6. After excluding those below 18 and non-German nationals, and applying the appropriate weighting variable (representation weight for Germany as a whole), this leaves 1994 cases; all analyses presented here only refer to voters (N = 1451), with Green voters (N = 170) compared with voters of other parties.

 7. The only panel data that exists comes from the German Socio-Economic Panel (SOEP). However, the measurement of Green support here is based on a question about party identification rather than voting behaviour. While the results are thus not necessarily comparable (for a discussion of the relative merits of either approach, cf. Klein and Arzheimer 1997, 1998, Kohler 1998a, 1998b), analyses based on SOEP data come to similar conclusions: Spiess and Kroh (2010) analysed Green support (i.e. identification) using panel data stretching from 1985 to 2007, with more than 14,000 individual cases. They also find that Green support is strongest among specific generations, in particular in what they term the ‘Schmidt generation’ who entered the electorate during Schmidt's chancellorship, i.e. those born between 1956 and 1964. Their findings have conflicting implications for the future of the Greens: they diagnose a high durability of the green project as carried by specific generations, but they also found that within these generations green support declines with age, suggesting longer-term problems for the Greens. A later analysis also using SOEP data (Kroh and Schupp 2011) suggests that the generation effect is much stronger than the age effect. Aging is associated with decreasing Green Party identification but at the same time, a substantial share of those voting Green in the 1980s maintain their commitment as they age. This means that the demographic change works in favour of the Greens. Combined with the Greens’ continued ability to appeal to younger age cohorts, the prognosis on the basis of this data is thus quite positive for the party.

 8. Klein and Arzheimer (1997) suggested that the move of the Greens to more moderate positions in the 1990s made them attractive to older voters that had not voted Green in the 1980s.

 9. This presentation was chosen as the most succinct way of modelling voting choice and follows the example of Clarke et al. 2004; the inclusion of multinomial model potentially raises the problem of the independence of irrelevant alternatives (IIA). For the models presented here, the Hausman text could not be performed as we need to use weighted data; and the Small-Hsiao test was mainly negative. However, Dow and Endersby (2004) as well as Kropko (2010) have shown that a violation of the IIA assumption does not invalidate the results. Furthermore, binary logistic regressions were conducted for each combination of parties, resulting in essentially identical results.

10. This statement of the share of the variance explained is based on the interpretation of the McKelvey and Zavoina's R2 statistic as the pseudo-R2 measurement most closely matching the R2 statistic in ordinary least squares (OLS) regressions (see Long and Freese 2006, p. 196).

11. The odds ratios displayed in Table 4 indicate changes in the odds of voting Green for each unit change in the independent variable, holding all other variables constant (cf. Long and Freese 2006, p. 178). The relative risk ratios reported for the multinomial regressions can be interpreted in a similar way (they are also defined as exp(b) and might be more appropriately called multinomial odds ratios).

12. While the number of cases in the GLES dataset is too small to model contextual influences, it was possible to compare West German and East German Green voters. All the regressions were also run for West and East Germany separately, with the interesting finding that regarding the environment as the first and second most important issue increases the odds of voting Green in East Germany by a factor of more than 10 (holding all other variables, as described in Table 5, constant). All other variables, with the exception of age and salience of nuclear energy, make no statistically significant difference. More than 20 years after unification, the image of the East German Greens as less left-wing and more environmentalist (cf. Hoffmann 1998) thus still applies.

13. This conclusion has to be qualified because some important alternative explanations could not be directly tested. Any future analysis should certainly include measurements of Inglehart's concept of post-materialist value change as well as a discussion of the theory of party identification and the factors that may, or may not, associate a stabilisation of Green voting with an increase in Green party identification.

14. The German Pirates Party was founded in 2006. It first attracted some attention during the 2009 federal election campaign with its radical civil rights agenda, in particular regarding internet content and access, and achieved a credible result with 2.0%. It drew support mainly from young male voters, attracting 12% of young men in the 18–24 age group (Infratest dimap 2009, p. 55). In the Berlin state elections on 18 September 2011, the Pirates polled 8.9% and entered the Berlin House of Deputies at the first attempt. Some commentators saw the Pirates, whose unconventional approach is frequently compared with the Greens of the early 1980s, as a potential threat to the Greens in the next federal elections (cf. Bergt and Schulze 2011). However, the demographic profile of Pirate voters – young men and unemployed (Forschungsgruppe Wahlen 2011) – indicates that the party is not particularly successful among the Green core constituencies of female voters in the '68 and NSM generations. An analysis of voter movements in Berlin found that the Greens only lost 17,000 of its 2009 voters to the Pirates (Bergt and Schulze 2011), less than 10% of the 2009 Green electorate and about 13% of all Pirates voters in 2011.