The exchange narrowed the differences down to a fairly small margin. They make it clear why a term like "excess deaths" is the only valid epidemiological one in view of the limitations on the historical evidence and the major part that famine played in the total.
Conquest challenges Davies's article in NLR 214.
He clarifies that his own calculation of XS deaths from 1930 to 1937 is of the order of 11 million. This figure is derived from the population deficit of 15 to 16 million in the suppressed census of 1937, by making a reasonable assumption about the division between deaths and a drop in births, he says.
He says a figure of 3 million more XS deaths in 1937-8 is open to dispute.
On executions, Davies accepts a figure of 681,000 1937-8, but Conquest appears to prefer a rough figure of 1.75 million given by the Head of Archives and by a representative of the Security Ministry.
Davies, in a reply, alleges that Conquest's pre-perestroika estimate totalled 17 million XS deaths, and implies Conquest has been forced to revise this downwards.
He claims that because we cannot know the true birth rate especially during the famine, the figure of 11 million XS deaths cannot be deduced from the 1927 and 1937 census data.
So the range of debate looks between 2/3 million and one and three-quarter million executions in 1937-8.
XS deaths, mainly from famine, 1930-8 range from 11 million to 17 million.
The Economist is presumably grubbing for headlines.
Chris Burford
London