![Sample our Politics & International Relations journals, sign in here to start your FREE access for 14 days]({"type":"image" , "src" : "/sda/5947/TOC-Subject-Sample-2021-Politics-International-Relations.jpg"})
Abstract
One of the most persistent findings in coalition research is the proportionality rule (‘Gamson's Law’) that guides the allocation of portfolios between parties. This article tests the assumption that a similar rule is at work within parties. More specifically, it examines the allocation of ministerial posts to regional party branches in 25 post-war cabinets in Austria between 1945 and 2008. Drawing on the literature on party organisations, three types of resources (membership, vote and population shares) are identified, all of which account for a substantial part of the variation in the shares of cabinet seats awarded to each party branch. The analysis thus bears out the proportionality proposition to a significant degree. Furthermore, it is shown that there is considerable cross-party variation in the allocation patterns, which reflects differences in the organisational structure of the parties.
Acknowledgements
This research has benefited considerably from the author's work under the auspices of the Austrian National Election Study (AUTNES), a National Research Network (NFN) sponsored by the Austrian Research Fund (FWF) (S10903-G11). An earlier version of this paper was presented at the workshop ‘Repräsentation regionaler Interessen in Parteien, Parlamenten und Regierungen’ [Representation of regional interests in parties, parliaments and governments] at the Dreiländertagung 2011, University of Basel, Switzerland. In addition to the participants of the workshop, the author would like to thank Marcelo Jenny, Thomas M. Meyer and Wolfgang C. Müller, as well as three anonymous reviewers for their valuable comments and suggestions.
Notes
Party factions represent another type of sub-unit that may play a significant role in intra-party portfolio allocation (see Leiserson, Citation1968; Mershon, Citation2001a, Citation2001b). However, as parties vary widely in their degree of factionalism, it is difficult to assemble consistent data and draw meaningful conclusions across factions and parties. The present analysis is therefore concerned exclusively with regional party branches.
Social Democratic Party of Austria (SPÖ); Austrian People's Party (ÖVP); the Freedom Party (FPÖ); Alliance for the Future of Austria (BZÖ); and the Communist Party (KPÖ).
In April 2005 the FPÖ's leadership split from the party and founded the Alliance for the Future of Austria (BZÖ). As of that time the Schüssel II cabinet was effectively a coalition between ÖVP and BZÖ. The analysis, however, always refers to the initial configuration of a cabinet.
The Freedom Parties in Vorarlberg and Upper Austria initially refrained from taking sides before eventually returning to the ranks of the FPÖ, though not without extracting concessions to alter the party statutes in their favour. In late 2009, the Carinthian Land party split from the BZÖ, renamed itself FPK (Freedom Party of Carinthia) and announced a close cooperation with the FPÖ in both the legislative and electoral arenas. For all practical purposes, the ‘old’ FPÖ has thus been restored in full organisational strength, leaving the BZÖ without any significant financial or human resources outside the national parliament.
However, all results of the subsequent analysis hold when controlling for Vienna in the regressions.
Owing to the fact that the number of seats at the cabinet table is limited, some Land parties are not awarded any portfolio at times. As the dependent variable thus contains a considerable number of zeros and hence violates the normality assumption, a simple OLS regression may not be the most appropriate estimation strategy. The OLS models are, however, reported in order to facilitate comparison with the classic portfolio allocation literature. As an alternative, Tobit models that can accommodate truncated dependent variables were estimated. The results do not suggest any substantive alteration of the conclusions drawn from the OLS models ().
Values range from r = 0.75 to r = 0.93.
The author would like to thank one of the anonymous reviewers for this suggestion.