The data used in the GCRI are open-source and widely accepted by the scientific community.
They are provided by institutions such as the World Bank, the UN, and a variety of academic institutions such as universities as well as research projects.
Data downloads for replication include data for EXCEL and .csv format (for STATA or R packages).
Version 5 comes with a new data management framework. All the data is now prepared as a single data frame, including all fully imputed variables, before the analysis process begins.
The files are available for download in a zipped archive.
Dataset alone, in .xlsx format.
GCRI data (v5.0.1, 0.7 MB).
Minor changes to the some variables.
GCRI data (v5.0, 0.7 MB).
Replication material, in .zip format.
Contains the dataset as well
as the data necessary to reconstruct it using included original
sources. Also included is a detailed table of the original sources, such
as original variable names and URLs.
Replication material (v5.0.1, 39.9 MB).
- Code cleaned
- Minor changes in some variables
Replication material (v5.0, 39.9 MB).
The data has been further processed to provide replication datasets suited for statistical processing. New in the February edition are:
Raw Data for Indicators (v3.0, 1.2 MB).
Raw data from 1989 to 2013, as found in the sources. The data was verified, referenced using country ISO3 code and systematically organized in Excel sheets. Missing data is represented by "x".
Normalized Data for Indicators (v3.0, 6.4 MB).
The raw data is normalized to a scale from 0 to 10, with 10 being the value assumed to contribute most to conflict risk. The Excel sheet is linked to the Raw Data. Missing data is represented by "x".
Normalized and Imputed Data for Indicators (v3.0, 6.4 MB).
The normalized data is imputed using a variety of techniques (see imputation methods (Word)). Imputed values are stored separately for transparency. This is the input data for the statistical models.
The model has been developed based on historical datasets of
dependent (output, i.e. conflict) and independent (input, i.e.
structural) variables. These are available in Excel.
Current conflict data (v2.0, 31 KB).
The most recent conflict data for all countries. This is
used to compare the GCRI 2015 output with the current state of conflict,
and derive differences, which may indicate areas of risk or
Current data (v2.0, 47 KB).
The most recent structural data for all variables and for
all countries. This is the input data used for to calculate the GCRI
Historical data (v2.0, 11 MB).
The 20 year historical time series used to derive the
statistical model. All independent and dependent variables are included.
Imputation steps are included.
Replication material for the December 2015 edition contains the data and scripts necessary to reproduce the regression and composite based scores. Documentation will be uploaded shortly (basic description of the data is available in the data replication file).
Replication material (.zip, 0,77MB)
Documentation (pdf, 2MB)
Smidt, M., L. Vernaccini, P. Hachemer, T. De Groeve; The Global Conflict Risk Index (GCRI): Manual for data management and product output; EUR 27908 EN; doi:10.2788/705817
This report describes the concept and methodology of the Global Conflict Risk Index (GCRI). It explains the scientific background, the statistical methods used, and the selection method for indicators. Tables, charts, and diagrams help to visualise our results and clarify our approach.
GCRI Concept and Methodology
De Groeve, T., P. Hachemer, L. Vernaccini; The Global Conflict Risk Index (GCRI): A Quantitative Model; EUR 36880 EN; doi:10.2788/184
The statistical models were derived using code written in the R language. The code and the data necessary to run it can be downloaded here: