Data

Job Insecurity Data

These data are measures of the job insecurity (likelihood of losing office) felt by leaders of the state. The data include a large cross-section of the globe from approximately 1950 to 2000. I estimate job insecurity following Cheibub (1998) as the likelihood that a leader loses office based on his/her time in office, cumulative number of previous leader changes, and economic growth. In a more complex estimation I use the above three covariates but also include the size of the leader's winning coalition, the number of previous irregular leader changes, and the age of the leader. In the data, the insecurity measures follow Cheibub's specification and the measures called "full" use the extended specification. To estimate these hazard rates, I use parametric survival models (Weibull). Additionally, I use the Archigos data from Goemans (2006) and BDM et al.'s data in separate measures to offer alternative indicators. Please cite Goemans (2006), Cheibub (1998), BDM et al. (2003), and Joseph Young. 2008. "Repression, Dissent, and the Onset of Civil War." PhD dissertation, Florida State University.
Stata

Protests Data

Episodes of Protest from Latin America, includes 18 countries, 124 episodes, and spans from 1990 to 2004. Excel | Stata

Do-files for model estimation and code for estimating multilevel models in MLwiN coming soon.

"What's Left for the Left in Latin America" Data

Data to replicate partisan effects on neoliberal reform in Latin America

Data | Model Replication Do-File | Graph Replication Do-File

US Counterinsurgency Operations in Iraq Data

Database of Coalition deaths and US COIN ops in Iraq from May 1, 2003 to Sept. 30, 2005
Coming Soon
Data | Do-Files | R Code | You need the PESTS program from Patrick Brandt to estimate

Useful Data Management Do-Files for TSCS Data

Do-File for creating country names from COW codes.
Do-File for creating Cow Codes from Country names. I adapted this from Jana von Stein's. Andy Beger has one as well.
Do-File for creating dates for state membership. If you need to impute data and do not want missing values for years when a state did not exist, use this fix (only for post 1945 world).