6 ideas
Plotting real income growth from 1988 to 2008 against each percentile of the global income distribution gives a shape like an elephant. The emerging Asian middle classes, around the global median, gained most; the global top 1% also gained heavily; the lower-middle classes of rich countries, around the 75th–90th percentiles, saw near-stagnation. Globalization's winners and losers are therefore specific positions in the world distribution, not whole nations.
The curve is anonymous: it compares the income at each percentile in 1988 with the income at that same percentile in 2008, but the people holding a percentile change as countries rise and fall. Growth-incidence curves describe how the distribution changed, not what happened to particular individuals.
Inequality does not rise once and then fall for good as economies develop, as the single inverted-U curve predicted. It moves in repeated waves: technological revolutions, structural change and globalization push it up, and forces such as war, disease, political pressure, education and low-skill-biased change bring it down. Rich countries are now on the rising side of a second wave driven by technology and globalization.
Before the twentieth century, falls in inequality came mostly from destructive events such as wars, epidemics and state collapse, not from benign policy. Twentieth-century equalization in rich countries also depended heavily on world wars and the political and fiscal changes they forced. This is a warning that today's rising inequality may not self-correct peacefully.
Most of the variation in income across all individuals in the world is explained by the country a person lives in, not by class or effort within that country. Being born in a rich country therefore gives an unearned rent, the citizenship premium, that outweighs most individual characteristics. This shifts global inequality from a question of class toward a question of place and makes migration a central equalizing force.
Global inequality treats the whole world as one population and ranks every individual by income, ignoring national borders. Seen this way, rising inequality within many countries sits alongside falling global inequality since roughly 2000, driven mainly by growth in China and India. The shift from class-based to location-based inequality over two centuries only becomes visible at this scale.