Gallagher Electoral Disproportionality Data
121 Countries, 1945-2014
Current version: v2
Data Updated: 20 March 2015
What is it?
This is a .csv formatted data set of the Least Squares Index of electoral disproportionality for 121 countries between 1945 and 2011. The Least Squares Index is also commonly referred to as the Gallagher Index. The data set combines Michael Gallagher's recently updated data (here) with some holes (see below) filled in with data made available by John Carey and Simon Hix (here).
Why did I make this?
There is currently no complete and recently updated Gallagher Index data freely available in a format that is easy to do statistical analysis on. The data on Michael Gallagher's website has been recently updated (the one I used was from 21 February 2012). However it is only available in PDF format; not exactly the easiest format to merge into an R or Stata file.
So, after going through the somewhat painstaking work of entering all of this data into a .csv file (and filling in some holes with Carey and Hix's data) I decided to make my efforts publicly available. Hopefully, this will free up some research time for more productive endeavours than copying and pasting data from a PDF file.
Import into R
To import the file directly into R simply use the
Data <- repmis::source_data(url = "http://bit.ly/Ss6zDO")
More details on the Gallagher Index
The Least Squares or Gallagher Index of electoral disproportionality basically uses the standard least squares method for comparing the relationship between two samples. It compares parties' votes to the legislative seats they are given. Higher numbers on the Gallagher Index indicate that there is a greater disparity between votes and seats, i.e. elections have produced more disproportional outcomes. A Gallagher score of 0 would indicated that the election produced perfectly proportional outcomes.
For more information see:
Gallagher, Michael. 1991. “Proportionality, Disproportionality, and Electoral Systems.” Electoral Studies 10(1): 33–41.
Carey, John M, and Simon Hix. 2011. “The Electoral Sweet Spot: Low-Magnitude Proportional Electoral Systems.” American Journal of Political Science 55(2): 383–397.
More details on the data.
These notes give specific details about how I combined the two data sets into the one .csv file.
Data is for national overall disproportionality.
I did not include Northern Ireland, Scotland, Wales, Turkish Cyprus, or Palestine.
All data is from Gallagher's PDF mentioned with the following exceptions.
Data for these countries was taken from Hix and Carey (see their notes for further details):
Guatemala before 2011
India before 2004
Indonesia for 1999
Mauritius before 2010
Papua New Guinea
South Korea before 2004
Sri Lanka before 2000
Thailand before 2011
Turkey before 2003
Ukraine before 2003
Venezuela before 2001
For years when there were two elections in one country I kept the latter one. This included:
Fiji uses September election for 1977 (previous election was in September)
France uses November election for 1946 (previous election was in June)
Greece uses November election for 1989 (previous election was in June)
Greece uses June election for 2012 (previous election was in May).
Iceland uses October election for 1959 (previous election was in June)
Ireland uses November Election for 1983 (previous election was in February)
Moldova uses July elections for 2009 (previous election was in April)
St Lucia uses April 30 for 1987 (previous election was in April 6)
United Kingdom uses October Election for 1974 (previous election was in February)
The United States data is for House elections. I did not included data on executive disproportionality, though Gallagher includes this in his PDF file.