Table 4

Mean inequality estimates, stratified by class of inequality measure and gender (equal weighting of observations)

Inequality measureAbsolute*Relative
person yearspersonsperson yearspersons
*Units of measurement for absolute measures are deaths per 100 000 (person years) and deaths per 10 000 (persons). †Absolute, the difference between the rates for the least and most advantaged socioeconomic groups; relative, the ratio of these rates. ‡Absolute, the decrease in the population rate that would occur if all groups had the rate of the most advantaged socioeconomic group; relative, this decrease as a proportion of the population rate. §Absolute, the number of deaths that must be “redistributed” to achieve equality; relative, this value as a proportion of the total number of deaths. ¶Absolute, the difference between the predicted rates for the most and least advantaged persons based on a weighted regression of socioeconomic group rates on socioeconomic status; relative, this difference as a proportion of the population rate. **Absolute, twice the area between the generalised concentration curve of health and socioeconomic status and the line of equality; relative, this value as a proportion of the population rate. ††Relative, an odds ratio generalised to ordinal data.
Rate difference/ratio†60923435171.771.591.831.77
Population attributable risk‡3111281570.
Index of dissimilarity§96.841.75.012.870.07660.07490.0780.0835
Slope index of inequality¶−829−345−54.2−41.6−0.64−0.62−2.36−3.67
Concentration indices**−0.00109−0.00046−0.00613−0.00332−0.086−0.083−0.091−0.096
Agresti’s ᆆNANANANA1.361.351.371.408