A note on the ‘natural rate of subjective inequality’ hypothesis and the approximate relationship between the Gini coefficient and the Atkinson index