Data@ECPR-METHODS: Government Composition and Partisanship Data, 1945-2014
This is an extension to the Woldendorp, Keman, and Budge dataset, which records government characteristics, cabinet composition, and partisan complexion. However, the Woldendorp et al. data end in the mid-1990s. Seki and Willaims continue their work extending the government composition data for 37 democracies through 2014. This includes the government composition data (characteristics of the government, type of government, tenure dates, etc) and the ministers (names, birth years, party affiliation, and genders) making up the cabinet.
New to this version of the data is a dataset of annual government (and PM) partisanship, calculated with both the manifesto data (MARPOR) and our own complexion of government and parliament (CPG) variables.
The latest release also includes a Stata do file that modifies the Seki-Williams government composition data in a number of ways. The unit of analysis can easily be changed so that it is the government-party (which is helpful for analysis of vote shares or party competition). The temporal dimension can also be modified to be daily, monthly, quarterly or annual.
Needless to say that you want to cite the data appropriately referring to the following publication:
Katsunori Seki and Laron K. Williams (2014). “Updating the Party Government Dataset,” Electoral Studies: 270-279.