`Statistics of Democide' has two purposes. First, it links all the relevant estimates, sources, and calculations for each of the case studies in Death by Government (`Demozid' - der befohlene Tod. Massenmorde im 20. Jahrhundert), and all additional cases of lesser democide for which data have been collected. The value of this is the listing of each source, its estimate, and comments qualifying the estimate. From these others can check and evaluate Rummel's totals, refine and correct them, and build on this comprehensive set of data. These data are presented and annotated for pre-20th century democide for the megamurderers and for the United States and lesser murderers. All data sources referenced in the democide tables are listed in the references. The methodological underpinnings for this collection have been given in Rummels previous work, i. e. Death by Government.
Second, having finished collecting all these data and completing the major case studies Rummel finally could systematically test the assumed inverse relationship between democracy and democide. That is the substance of this book. Rummel details the tests and summarizes them.
Conclusion is that the diverse tests are positive and robust, that the less liberal democracy and the more totalitarian a regime, the more likely it will commit democide. The closer to absolute power, the more a regime's disposition to murder one's subjects or foreigners multiplies. As far as this work is concerned, Rummel concludes: "it is empirically true that Power kills, absolute Power kills absolutely."
R. J. Rummel is Professor Emeritus of Political Science, University of Hawaii.