Gini's mean difference and the measurement of absolute mobility

Three new indices of income mobility are proposed based both on an absolute and a "welfarist" approach to mobility measurement. They all have a "growth" and a "structural mobility" component. In addition the second and third indices have also an "exchange mobility" component, the second one assuming that re-ranking has a positive impact on welfare, the third one taking the opposite view. All three indices increase when an equal amount of money is added to all incomes or when all incomes are multiplied by a constant, assuming this constant is greater than one. Mobility increases also when money is transferred from a richer to a poorer individual. All three mobility indices obey Dalton's principle of population replication. Finally an income swap between two individuals has a positive impact on mobility, as far as the second and third indices are concerned, and this impact is stronger, the greater the income and rank difference between these two individuals. The empirical illustration based on Israeli data showed that income mobility is generally higher among men than women. It tends also to be higher, the higher the educational level. Finally mobility seems to be highest among Jews born in Israel and lowest among those born in Asia or Africa, the level of income mobility among Jews born in Europe or America being between that of the two other groups.