A new perspective on inequality: The income distribution across advanced countries
By Mikkel Hermansen, Economist, OECD Economics Department
What is the relevant perspective for evaluating people’s living standards in advanced countries? According to standard assessments of inequality it is fellow citizens within the country. In a recent paper (Hermansen, 2017), I argue that an interpersonal income distribution across advanced countries can provide a useful complement to comparisons of relative inequality across countries. For instance, the United States is well-known to be more unequal than Sweden, but how do the incomes of the bottom and top of the distributions across these and other countries compare directly in PPP terms?
Figure 1 illustrates how countries (grouped by regions or similar welfare models) are spread across the income distribution for advanced countries. Because this is a homogeneous set of countries, almost all subgroups are represented across the distribution from the bottom 5% to the top 5%. The United States stands out as it composes more than half of the top 5% of the aggregate distribution, but also makes up more than 30% of the bottom 5%. By contrast, the egalitarian Nordic countries are concentrated in the upper half, but nearly absent from the tails.
Inequality across all individuals living in advanced countries reaches 37 based on the Gini index, which has a scale going from 0 (perfect equality) to 100 (all income accruing to one person). This is relatively high, corresponding to the inequality levels observed in Israel and Estonia, but lower than inequality within the United States.
Inequality across advanced countries has increased by almost 3 Gini points from the mid-1990s to 2013, roughly in two phases, first from the mid-1990s to the mid-2000s and then again in the aftermath of the crisis. This was mainly driven by the top and bottom 10% drifting apart, as can be seen from the growth incidence curve in Figure 2. This curve relates incomes across the aggregate distribution for advanced countries in 2013 to the same percentile in 1995, ignoring changes in country composition across the distribution (as illustrated in Figure 1).
The growth incidence curve is a powerful and popular tool in inequality studies since it can provide granular information not available from measures such as the Gini. In the paper, I present a new approach to decompose the growth incidence curve into a within- and between-country component, allowing further granularity of within- and between-country inequality changes. For advanced countries, the between-country curve is almost flat from 1995 to 2013, meaning little change in inequality when replacing all households’ income with average income of their country of residence. This implies that almost all of the rise in inequality across advanced countries is driven by the well-known rise in inequality within countries, reflected by the parallel shape of the within-country component and the growth incidence curve in Figure 2.
Hermansen, M. (2017), “The global income distribution for high-income countries”, OECD Economics Department Working Papers, No. 1402, OECD Publishing, Paris, http://dx.doi.org/10.1787/65206dc1-en.