MPIDR Working Paper

Gini coefficient as a life table function: computation from discrete data, decomposition of differences and empirical examples

Shkolnikov, V. M., Andreev, E. M., Begun, A. Z.

MPIDR Working Paper WP-2001-017 (June 2001).
Rostock, Max Planck Institute for Demographic Research

Revised September 2004 / Also published as: MPIDR Working Paper WP-2006-047


This paper presents a toolkit for measuring and analysing inter-individual inequality in length of life by Gini coefficient. Gini coefficient is treated as an additional function of the life table. A new method for the estimation of Gini coefficient from life table data has been developed and tested on the basis of hundreds of life tables. The method provides precise estimates of Gini coefficient for abridged life tables even if the last age group is 85+. New formulae have been derived for the decomposition of differences in Gini coefficient by age and cause of death. A method for further decomposition of age-components into effects of mortality and population group has been developed. It permits the linking of inter-individual inequalities in length of life with inter-group inequalities. Empirical examples include the decomposition of secular decrease in Gini coefficient in the USA by age, decomposition of the difference in Gini coefficient between the UK and the USA by age and cause of death, temporal changes in the effects of elimination of causes of death on Gini coefficient, and decomposition of changes in Gini coefficient in Russia by age and educational group. Consideration of the variations in Gini coefficient during the last decades and across modern populations show that these variations are driven not only by historical shifts in the distribution of deaths by age, but also by peculiar health and social situations. (AUTHORS)