A data set on the economic determinants of political unrest, for replicating a publication from 1973.
Format
A data frame with 26 observations on the following 9 variables.
countrya character vector for a country name
levviola numeric vector for the level of violence
pcia numeric vector for per capita income
incdista numeric vector for income distribution
d_pcia numeric vector for per capita income growth
sema numeric vector for socioeconomic mobility
cominta numeric vector for communication intensity
concfaca numeric vector for concentration factor
popa numeric vector for population size
Details
The bulk of these data come from Russett's (1964) World Handbook of Political and Social Indicators. The data themselves are transcribed from the appendix of the article, which allows a replication of the results that Parvin (1973) reports. You should read that article for more information as to what's happening and for what purpose.
I did not catch Parvin (1973) mentioning this in the article, but there must be some kind of additive constant in the level of violence variable because the logarithmic transformations he reports would be undefined for the observations (like Denmark) where the level of violence is zero. The easiest way to approximate whatever Parvin (1973) did is to add .001 to the level of violence variable before taking its logarithmic transformation. That would allow a near perfect replication of Table 1.
It should go without saying that the population reported for Belgium, in the appendix, is likely a transcription error. Belgium's population is reported here as 9184, not "91.84.00".
The United Arab Republic was the short-lived union of Egypt and Syria, if you were curious what that is in the data.