Disentangling the Value of a Ministry: Party Leaders’ Evaluations of German State Ministries

Abstract

While the role played by ministries in the process of coalition government has been investigated from multiple angles, there is a clear lack of knowledge about which specific ministry features party leaders actually value as they assess different government posts. This paper aims at discovering whether, next to office considerations, the policy influence resting with a ministry does affect its value. A new survey of party leaders in the German states enables us to estimate the relative importance of specific office (e.g. public standing) and policy considerations (e.g. influence via legislation) for ministries’ values as well as to directly investigate differences between parties. The results show that both office and policy considerations matter for ministry evaluation generally, but also that different aspects have different weights. Furthermore, while all parties value the policy influence of a ministry, there is variation as to which type of policy influence (cross-sectional vs. within-jurisdictional) parties emphasise.

Acknowledgements

This manuscript was written in the context of the research project ‘Government Formation as an Optimal Combination of the Office and Policy-Motivation of Parties’. We are grateful to the German Research Foundation (DFG) for its financial support. For institutional support, we thank the Hanse-Wissenschaftskolleg/Institute for Advanced Study in Delmenhorst. We further thank Franz U. Pappi and Jana Windwehr who participated in the development and preparation of the questionnaire on which this paper’s analyses are based. Finally, we would like to thank the members of the standing group ‘Handlungs- und Entscheidungstheorie’ (German Political Science Association), Christian Martin and two anonymous reviewers for helpful comments.

Notes

1. Green parties stood out as weighting policy considerations heavily. As in the past these parties have usually been niche parties focusing on one particular policy field (see Adams et al. 2006; Meyer and Miller 2013), policy considerations should clearly drive their evaluation of ministries. Yet, since we are investigating party leaders in contemporary Germany and the Green parties in the German states hardly qualify as niche parties any longer, we suggest investigating this potential contingency in future analyses.

2. Druckman and Warwick (2005) use a similar strategy of providing respondents with a benchmark rating.

3. The policy fields are: finance; interior; justice; culture and education; Federal and European matters; economy and transport; labour and social affairs; agriculture; construction; environment and planning (see Bäck et al. 2011; Pappi et al. 2008). If a ministry, for example, consists of three departments which are concerned with the policy field of education and of one additional department concerned with social affairs, the respective overall saliency of the ministry is 0.75 times the saliency of culture and education plus 0.25 times the saliency of labour and social affairs.

4. See Linhart and Windwehr (2012) for an extensive overview on the generation of these data.

5. These summary results have to be read with the caveat in mind that some departments are usually situated in the same ministry (e.g. Economy, Transport and Construction) which prohibits disentangling which department is driving the overall ministry rating. However, Linhart and Windwehr (2012: 595) have focused on specific policy fields and arrive at very similar results regarding the importance of particular departments. Extensive descriptive statistics for the dataset are available on the following website: http://www.mzes.uni-mannheim.de/publications/misc/linhart/2012zparl_Daten.xls. The questionnaire (in German, example from Hessen) is available at http://www.mzes.uni-mannheim.de/publications/misc/linhart/2012zparl_Fragebogen.pdf.

6. There is also a general similarity between the German state level (rated by leaders) and the federal level (rated by experts; see Druckman and Warwick 2005: 39). On both levels, the prime minister is followed by Finance; Interior and Economy each hold relatively high positions. In each case, following these top posts, many different ministries are receiving very similar ratings. One clear difference, however, is that the gap between the head of government and the ministries is considerably larger on the federal level where the head of government is seen as more than twice as valuable as the average ministry. Other differences reflect differences in competencies between the federal and the state level: Education is rated much higher on the state level since here state governments enjoy relatively greater levels of autonomy from the federal government (see Kropp and Sturm 1998: 33). For Justice this division of competencies is reversed (see Prätorius 2006), thus explaining Justice’s low rating in our party leader survey.

7. The coefficients for the state dummies are not reported in the regression tables. Regarding the coefficients of the state dummies, Bavaria and Rhineland-Palatinate differ significantly from the state of Brandenburg (base state) in that ministries are rated higher in these states. The constant reported for the different models is that for the base state Brandenburg. Full results (including full information models where all three ranks are considered and state dummy coefficients are reported) are provided in an appendix.

8. This is confirmed in the results for Model 5 where, additional to the policy considerations, office considerations are taken into account. In this model, manifesto saliency remains a significant predictor of a ministry rating, as the further policy variables do. Furthermore, as one of our reviewers pointed out, the saliency of a ministry might moderate the effect that this ministry’s policy influence has on its rating. Initial testing does not confirm this, but we suggest that this potential contingency is investigated in-depth by future research.

9. This is also confirmed when considering the full information about different rankings in the model (see Appendix). Being ranked as the most important ministry regarding cross-sectional responsibility (usually this rank will be held by the Ministry of Finance) has a very strong effect on a ministry’s rating while being ranked as second or third most important does not exert a significant effect.

10. The models again include state dummies, the coefficients and standard errors of which are not reported.