1992 United Kingdom electoral returns
Electoral returns, selected constituencies, 1992 general election for the British House of Commons
data(UKHouseOfCommons)
A data frame with 521 observations on the following 12 variables.
constituencya character vector, name of the House of Commons constituency
countya character vector, county of the House of Commons constituency
y1a numeric vector, log-odds of Conservative to LibDem vote share
y2a numeric vector, log-odds of Labor to LibDem vote share
y1laga numeric vector, y1 from previous election
y2laga numeric vector, y2 from previous election
coninca numeric vector, 1 if the incumbent is a Conservative, 0 otherwise
labinca numeric vector, 1 if the incumbent is from the Labor Party, 0 otherwise
libinca numeric vector, 1 if the incumbent is from the LibDems, 0 otherwise
v1a numeric vector, Conservative vote share (proportion of 3 party vote)
v2a numeric vector, Labor vote share (proportion of 3 party vote)
v3a numeric vector, LibDem vote share (proportion of 3 party vote)
These data span only 521 of the 621 seats in the House of Commons at the time of 1992 election. Seats missing either a Conservative, Labor, or a LibDem candidate appear to have been dropped.
The original Katz and King data set does not have case labels. I used matches to an additional data source to recover a set of constituency labels for these data; labels could not recovered for two of the constituencies.
Jonathan Katz; Gary King. 1999. "Replication data for: A Statistical Model of Multiparty Electoral Data", http://hdl.handle.net/1902.1/QIGTWZYTLZ
Katz, Jonathan and Gary King. 1999. “A Statistical Model for Multiparty Electoral Data”. American Political Science Review. 93(1): 15-32.
Jackman, Simon. 2009. Bayesian Analysis for the Social Sciences. Wiley: Chichester. Example 6.9.
data(UKHouseOfCommons)
tmp <- UKHouseOfCommons[,c("v1","v2","v3")]
summary(apply(tmp,1,sum))
col <- rep("black",dim(tmp)[1])
col[UKHouseOfCommons$coninc==1] <- "blue"
col[UKHouseOfCommons$labinc==1] <- "red"
col[UKHouseOfCommons$libinc==1] <- "orange"
library(vcd)
vcd::ternaryplot(tmp,
dimnames=c("Cons","Lab","Lib-Dem"),
labels="outside",
col=col,
pch=1,
main="1992 UK House of Commons Election",
cex=.75)Please choose more modern alternatives, such as Google Chrome or Mozilla Firefox.