A data set on domestic conflict events in 2012 as recorded by the Cross-National Time Series Database. Data exist for teaching about count models.
Format
A data frame with 198 observations on the following 19 variables.
iso2ca two-character ISO code
countrya character name for the country corresponding with the ISO code
assassinationsthe count of assassinations in 2012
strikesthe count of general strikes in 2012
guerwarthe count of guerilla warfare events in 2012
govtcrisesthe count of government crises in 2012
purgesthe count of purges in 2012
riotsthe count of riots in 2012
revolutionsthe count of revolutions in 2012
agdthe count of anti-government demonstrations in 2012
wcithe weighted conflict index in 2012
areathe land area in square kilometers
adultpopthe adult (15+) population (in 1000s)
youthpopthe youth (15-29) population (in 1000s)
gdppcGDP per capita (in constant 2015 USD)
urbanshareurban population over total population (as percentage)
tpoptotal population (in 1000s)
polyarchyelectoral democracy index, an estimate of democracy
perctserpercentage of tertiary school-aged population enrolled in tertiary school
Details
Conflict events data come from the Cross-National Time Series Database. I've used these data before for published papers, but the relative opacity of a data set for yearly purchase comes with a bit of a caveat emptor for the important question of real-world inference.
Data on the democracy estimate and tertiary school enrollment rate come from the Varieties of Democracy project. Democracy estimate for Palestine comes as a simple average of the two Palestinian territories collected by the Varieties of Democracy project. These are West Bank and Gaza. The tertiary school enrollment variable, which originally comes from a data project by Barro and Lee (2013), is "filled" to the referent year from the most recent year available in the data. That would be 2010. It's fine for this purpose.
Population estimates come from the UN Population Division. GDP per capita comes from the World Bank. The estimate of land area (in square kilometers) comes from the CNTS. Country name comes from CNTS as well.
In all but the case of the data from CNTS, and the "filled" case of the tertiary school enrollment variable, the referent year for the data is 2011. Not that anyone is going to care too much for a simple data set like this, but this would be the ol' endogeneity concern.