As incomes in the bottom half of populations grow, does inequality rise or fall?

Which nations have most raised the incomes of the bottom 50% of their populations? Which have failed to do so?

2000:2019 When the bottom 50% rise, so too inequality

Simply as a matter of logic, when incomes increase or decrease at the bottom of the distribution, the gap between rich and poor has no pre-determined trajectory. Inequality could increase when the poor become richer, or the opposite, inequality could decline. The figure shows, however, that, averaging across all the world, when the poor in an economy become better off, inequality there rises. (All the calculations here are mine and build on the data in the World Inequality Database.)

The non-parametric local regression line – a locally estimated scatterplot smoothing or LOESS line, in blue – shows a consistent upward slope. The shaded area surrounding that line is the 95% confidence interval on the estimate. In the figure the horizontal axis measures how many times average incomes of the bottom 50% of the population (in USD) rose between 2000 and 2019. A value less than 1 indicates that average income fell. The vertical axis is the same ratio for inequality, measured as the USD difference between average incomes of the top 10% and bottom 50%.

In the figure, China led the world both in raising the incomes of its bottom 50% (over four-fold) and in inequality increase (almost six-fold). In Singapore the bottom half improved average incomes two-fold, about the middle of the collection of all the world’s economies, but towards the top when compared to other developed countries.

Norway, often held up as an example of egalitarian practice, saw its bottom half increase average income by 52% and its inequality rise 50%. Compared to the local regression line, Norway’s inequality increase was just about average given the income improvement of its bottom 50%.

After a fall in their average income over the first decade, the US’s bottom 50% managed an increase of 12% between 2000-2019. In this the US was close to many other economies – those to the left of 1 on the horizontal axis – that actually saw an immiserisation of their already poor, i.e., the bottom half of the population suffering a decline in average income.

But income and mobility experiences are diverse. Both Thailand and Armenia increased over three-fold the average incomes of the bottom halves of their respective populations. But Thailand did this with a much lower increase in inequality than did Armenia.

2000:2019 Residual inequality against bottom 50% growth

The second figure plots residual inequality against the same horizontal axis, namely improvement in that economy’s poor. By residual inequality I mean how much inequality changes differ from average (the fitted local regression line). This second figure is, therefore, just a sharpening of the information in the first figure.

In this figure those economies above the 0 horizontal came in with inequality changes higher than average. Again, notice Norway is just about average, Singapore a bit above average, Thailand, below average.

The Supply Curve for Inequality

A direct and natural interpretation of Fig. 1 draws on the common aphorism “a rising tide lifts all boats”: economic growth overall raises incomes of both rich and poor. However, it does so more for those already well off than for those at the bottom of the income distribution.

A second view draws on so-called “trickle-down” thinking, where the economy is organised in such a way that gains enjoyed by the rich spill over, in diminished form, to the poor. Hence, inequality rises, but the poor are lifted in the process.

These two views differ in their hypothesis on the primary causal drivers of income distribution dynamics. Is there an external force, the “rising tide”, exogenous to the incomes of both rich and poor that shifts both the latter simultaneously? Or is the exogenous engine actually rich people’s acquiring greater resources that then trickle down, endogenously, to the poor?

A third distinct interpretation is that the Fig. traces out a supply curve. That supply schedule addresses the question, What price, measured in increases in inequality, do societies have to pay if they seek to raise the incomes of the bottom 50% of their populations? How costly is lifting the poor? For a given quantum of social mobility, society has to pay a price in potentially having to tolerate the rich becoming even richer, i.e., rising inequality. Conversely, of course, inequality might well fall with social mobility. But whether inequality rises or falls, and how large that tradeoff is can only be ascertained from empirical evidence. Fig. 1 provides that evidence.

(Excerpt from my working paper, Inequality is Not a Sufficient Statistic.)

  1. What are your sources for China’s stats?

    The current World Bank Gini figures put the US at 41.8 and rising; China’s at 38.1 and falling.
    See https://data.worldbank.org/indicator/SI.POV.GINI

    Xi’s plan for 2021-2035 is to bring Chinese Gini below Finland’s current 27.2.

    Incidentally, thanks to 99% home ownership, Wealth Gini indices look like this:

    China 0.547
    France 0.730
    USA 0.801
    https://wid.world
    https://en.wikipedia.org/wiki/List_of_countries_by_distribution_of_wealth

    • Danny Quah

      You’re spot on, obviously, with the Ginis. They capture critical trends. However, I use, instead, the separation between the average incomes of the top 10% and bottom 50% as my measure of inequality, ineqQ. I calculate these average incomes from the income shares data provided in https://wid.world; these statistics are not themselves directly found in that database. I find it easier to communicate to non-specialists what these inequality indicators measure.

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.

%d bloggers like this: