Summary
The King-Hardy method and the Hardy method mentioned in e.g. Miller (1946, 6.4) are old, well known moment methods in Gompertz-Makeham graduation of mortality. They have poor efficiencies (Forsén, 1977, App. F.6 and F.5), measured in terms of asymptotic variance, relative to the modified minimum chi-square method which is optimal (Hoem, 1972, 6.2, 7.2). This paper shows that a modification of the moment methods mentioned, gives a method that is “almost” as efficient as the best method available, and it is easier to use in practice.