![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
Recent research suggests that committees in parliamentary democracies may, at least partly, be endogenous to the prevalence of coalition government. In this article, I examine the conditions under which parliamentary majorities reform legislative rules to expand or reduce committee power. I expect that, ceteris paribus, the greater the conflict inside the governing coalition, the higher the probability that parties in government will adopt reforms expanding committee power and the lower the chance that they will implement changes reducing such power. These expectations are tested using original new data on the reforms of committee agenda powers undertaken in eight European states within 20 years from democratic transition. I find some evidence to support the endogeneity of committee power to the ideological heterogeneity of parliamentary government.
Acknowledgements
An earlier version of this paper was presented at the Second Conference of the ECPR Standing Group on Parliaments in Vienna, 26–28 June 2014, and at the Politics Research Colloquium held on 20 October 2014 in the Department of Politics and international Relations at Oxford University. I would like thank Nancy Bermeo, Elias Dinas, Dave Doyle, Andy Eggers, Marcelo Jenny, Shane Martin, Heike Klüver, Andreas Murr, Sergi Pardos-Prado, Gwen Sasse, and two anonymous reviewers for useful comments and suggestions. Special thanks to Agnė Ambrazevičiūtė, Irena Bačlija, Mátyás Bencze, Petr Bouchal, Katarína Chovancová, Agnieszka Cieleń, Niels Goet, Raphael Heuwieser, Jessie Hronesova, Riho Kangur, Ágnes Kovács, Matevz Malcic, Monika Marekova, Karina Oborune, Marta Riksa, Rodana Rozanc, Joonas Taras, Petra Weingerl, Vilija Velyvyte and Zsófia Zsoldics for research assistance.
Disclosure Statement
No potential conflict of interest was reported by the author.
Notes
1. This is a boundary condition for my expectations which, however, is difficult to test empirically, as committee power may never take maximum or minimum values.
2. This causal mechanism does not directly follow from Martin and Vanberg’s model, but is consistent with their theory.
3. The EU accession of Romania and Bulgaria has followed a slightly different temporal path and these countries are excluded from this study.
4. The dataset is available from http://www.erdda.se.
5. I excluded: Peterle, Drnovsek (Slovenia), Fischer, Necas (Czech Republic), Kubilius (Lithuania), Orban II (Hungary), and I added Pithart (Czech Republic), Meciar 0, Carnogursky (Slovakia). The data on the three cabinets not included in the ERD dataset come from http://www.parlgov.org.
6. These cabinets are: Siimann in Estonia; Horn in Hungary; Drnovsek II and Drnovsek VI in Slovenia.
7. The four proxies correlate at between 0.58 and 0.98 (significant at the 0.001 level).
8. The effective number of cabinet parties is calculated as one divided by a sum of squares of each party’s proportion of all government seats.
9. The weighted cabinet range is calculated using RILE party scores as a square root of the sum of products of each government party’s share of cabinet seats and the squared difference between the party position and the size-weighted cabinet mean position.
10. An analysis of the data using ordinal logit regression produces similar results.
11. The maximum number of observations is 297. It declines to 261 in models 1 and 2 (Table ) because no power expansions occurred in Slovakia and hence these cases are removed from the analysis. The number of observations declines to 246 in models 3 and 4 (Table ) and to 282 in models 7 and 8 (Table ) due to missing CMP data for parties in Latvia.
12. A closer inspection reveals that the low statistical significance of these results is due to an outlier cabinet (Drnovsek II in Slovenia) which, although highly ideologically heterogeneous, implemented reforms decreasing committee agenda powers. Once this observation is removed from the analysis, both coefficients reach standard levels of statistical significance.
13. For the purposes of this robustness check, I assume equal weight of each power dimension. Accordingly, I sum reform scores (coded as –1, 0 or 1) across the three dimensions and if the sum is positive I assign a score of 1, if it is negative, a score of –1, and 0 otherwise. In principle, it would be possible to create an ordinal variable ranging from –3 to +3, but in practice there are very few observations in the top categories at each end of the spectrum. Note also that under this aggregation procedure reform scores may cancel each other out.