Counting Soviet deaths in the Great Patriotic War : comment

Michael Haynes has pointed out that the conventional methodology for estimating the number of Soviet war deaths provides only a lower limit. It calculates wartime “excess” deaths, assuming that no one was killed by the war who would normally have died anyway; this sets a lower bound on the number of real war deaths that may have resulted directly or indirectly from enemy action. Where is the upper limit? Haynes proposes the 16 million total of “normal” deaths as a measure of the maximum possible downward error when real Soviet war deaths are estimated by the excess mortality method. The possible margin of downward error arising from this method can be calculated and is not 16 million but approximately 1.9 million