Religious Characteristics of States Dataset Project - Demographics v. 2.0 (RCS-Dem 2.0), COUNTRIES ONLY

DOI

10.17605/OSF.IO/7SR4M

Citation

Brown, D., & James, P. (2019, February 10). Religious Characteristics of States Dataset Project - Demographics v. 2.0 (RCS-Dem 2.0), COUNTRIES ONLY.

Summary

The RCS-Dem dataset reports estimates of religious demographics, both country by country and region by region. RCS was created to fulfill the unmet need for a dataset on the religious dimensions of countries of the world, with the state-year as the unit of observation. It covers 220 independent states, 26 selected substate entities, and 41 geographically separated dependencies, for every year from 2015 back to 1900 and often 1800 (more than 42,000 state-years). It estimates populations and percentages of adherents of 100 religious denominations including second level subdivisions within Christianity and Islam, along with several complex categories such as "Western Christianity." RCS is designed for easy merger with datasets of the Correlates of War and Polity projects, datasets by the United Nations, the Religion And State datasets by Jonathan Fox, and the ARDA national profiles.

Data File

Cases: 42747
Variables: 225
Weight Variable: None

Data Collection

October 2013 to December 2014, October 2016 to May 2017

Original Survey (Instrument)

RCS Demographics Codebook v2.0

Funded By

The Association of Religion Data Archives

Collection Procedures

RCS-Dem pools estimates from several dozen present-day and historical sources, combined, interpolated, and extrapolated to form 42,747 state-year estimates of 100 denominations each. For the precise territorial and temporal coverage of each political entity, please see the codebook section Appendix A in the RCS Demographics Codebook above.

Four countries are double-counted due to ambiguities in their sovereign status or in which other state they belong: Bosnia and Herzegovina, Egypt, Syria, and the West Bank. A table detailing the time period of and rationale for double-counting them is provided at the end of Appendix A in the codebook available for download.
Country-level estimates are combined to form region-level estimates for 30 regions and sub-regions demarcated by the United Nations Statistics Division, and the entire world. For the precise territorial and temporal coverage of each region, please see the codebook section Appendix B in the RCS Demographics Codebook above.

A taxonomy of religions covered in RCS, along with their numerical identifiers, is provided at Appendix C.

The source of each population or percentage estimate, or the method by which it is derived, is documented for every data point. This information is not provided in the final release due to considerations of space; the documentation is contained within several hundred Excel worksheets with a total size of 2.5 Gigabytes. The working files will be made available by the author upon request and supply of storage media. Additional details from the primary researcher discussing population estimates sources population or percentage within each religious denomination are found in the codebook.

Intercoder reliability checks were performed on estimates of populations that were (1) obtained from multi-volume almanacs, or (2) derived by collapsing narrower categories of churches into broader subdivisions of Christianity. Tests on five percent random samples were performed on data obtained from the following sources: the CIA World Factbook, the Kettani article series, the World Churches Handbook, and the World Christian Encyclopedia. The Intercoder reliability rate was 85 percent or higher in all four tests.

Sampling Procedures

Direct observations of populations and percentages in selected country-years are obtained primarily from official government sources and 25 secondary sources (all of which are listed in the Codebook). Each direct source is documented and the methodology for selecting sources is outlined in the Codebook.

Estimates between direct observations are interpolated primarily via exponential curves, polynomial curves, and linear increments of percentages. Estimates beyond first or last direct observations are extrapolated in a similar manner. Guidelines for selecting methodologies are outlined in the Codebook.

Principal Investigators

Dr. Davis Brown, Nonresident Fellow, Baylor University Institute for Studies of Religion
Dr. Patrick James, Dornsife Dean's Professor of International Relations, University of Southern California

Related Publications

An article of record is due for publication in 2018.