One-Thirteenth of a Data Point Does Not a Generalization Make: A Response to Dulić

While Dulić’s article is a helpful assessment of the sources Rummel used, it fails as a test of his general democide estimation methods or as an evaluation of his democide estimate for Tito’s Yugoslavia. Aside from mistakenly generalizing from one regime to Rummel’s results for 218, Dulić ignores 39 years of Rummel’s estimates for 1949–87 to concentrate on his own estimation of what Rummel’s estimate would be for 1944–48 (12.8% of the period Rummel covered). Dulić loads his estimate with war killed and the greater democide of other factions for 1941–45, miscompares the total to the 1948 census, and thereby wrongly concludes that Rummel’s estimate for the full 1944–87 period is too high. But Dulić provides no comparative estimate of his own for the full period to show this.