Cross-national data on relative size of the trade unions and predictors in 20 countries. This is a data set of interest to replicating Western and Jackman (1994), who themselves were addressing a debate between Wallerstein and Stephens on which of two highly correlated predictors explains trade union density.
uniondensityA data frame with 20 observations on the following 5 variables.
countrya character vector for the country
uniona numeric vector for the percentage of the total number of wage and salary earners plus the unemployed who are union members, measured between 1975 and 1980, with most of the data drawn from 1979.
lefta numeric vector tapping the extent to which parties of the left have controlled governments since 1919, due to Wilensky (1981).
sizea numeric vector measuring the log of labor force size, defined as the number of wage and salary earners, plus the unemployed.
concena numeric vector measuring the percentage of employment, shipments, or production accounted for by the four largest enterprises in a particular industry, averaged over industries (with weights proportional to the size of the industry) and the resulting measure is normalized such that the United States scores a 1.0, and is due to Pryor (1973). Some of the scores on this variable are imputed using procedures described in Stephens and Wallerstein (1991, 945).
Data documentation are derived from Simon Jackman's pscl package.
I just tidied up the presentation a bit.
Jackman, Simon. 2009. Bayesian Analysis for the Social Sciences. Wiley: Hoboken, New Jersey.