Published online by Cambridge University Press: 13 October 2014
This article considers the historical development of a characteristic crucial for the functioning and normative appeal of Westminster systems: cohesive legislative parties. It gathers the universe of the 20,000 parliamentary divisions that took place between 1836 and 1910 in the British House of Commons, construct a voting record for every Member of Parliament (MP) serving during this time, and conducts analysis that aims to both describe and explain the development of cohesive party voting. In line with previous work, it shows that – with the exception of a chaotic period in the 1840s and 1850s – median discipline was always high and increased throughout the century. The study uses novel methods to demonstrate that much of the rise in cohesion results from the elimination of a rebellious ‘left tail’ from the 1860s onwards, rather than central tendency shifts. In explaining the aggregate trends, the article uses panel data techniques and notes that there is scant evidence for ‘replacement’ explanations that involve new members behaving in more disciplined ways than those leaving the chamber. It offers evidence that more loyal MPs were more likely to obtain ministerial posts, and speculates that this and other ‘inducement’-based accounts offer more promising explanations of increasingly cohesive parties.
Associate Professor, Department of Politics and International Relations, and Professorial Fellow, Nuffield College, University of Oxford (email: [email protected]); Associate Professor, Department of Government, Harvard University (email: [email protected]). Kathy Wang, Tania Amarillas and Jack Blumenau provided excellent research assistance. We are grateful to the History of Parliament Trust for data. We also gratefully acknowledge the financial support of STICERD and IQSS. Michael Gill, J.F. Godbout, Peter Hall, Brandon Stewart, audiences at the IQSS Applied Statistics Seminar, the Australian National University and Midwest Political Science Association provided helpful comments on an earlier draft. All data used for this article will be made available on the DVN network on publication. Data replication sets and online appendices are available at http://dx.doi.org/doi: 10.1017/S0007123414000362.