1.1. Income Inequality: Meaning and Measurement

1.1.1. How is Income Defined?

First, income has to be comparable from one perspective or another: over time, across geographic space, or among households, just to name a few. In other terms, income needs to be expressed in one single unit of measurement, whether it refers to the choice of a particular currency for international comparisons, the base year of a price index for temporal comparisons, or the percentage of a total if one considers income shares. Furthermore, the unit of account may be selected carefully. Champernowne and Cowell (1998, p. 68) illustrate the point with the following example.

‘We cannot assume that an income of x ₁ going to ‘person’ 1 always counts as less than income of x ₂ going to ‘person’ 2 just because x ₂ is a larger number than x ₁, since ‘person’ 1 might be a single individual and ‘person’ 2 a sprawling household of three adults and ten children, but incomes must be defined in such a way that we can always decide for any pair which is the larger or that they are equally large, even when the two ‘persons’ receiving them are demographically dissimilar: that is what is meant by the income being comparable.’

Second, one has to be clear about the income sources counted in the definition of income, as it varies from one statistical agency to the other, or from one year to another within the same agency. Income usually includes wages, business and property incomes, dividends, interest, and social benefits. But official income may exclude amounts that contribute to the individual’s overall living standard, such as transfer payments, non-cash income (the use of company car, subsidized meal vouchers, food stamps), and so on. In other words, the income definition needs to be clearly stated. The income definition used here is discussed in section 3.3.2 of this thesis.