Graphic realism: Jason Hickel shows that, when you measure global wealth gaps, how you visualise really matters


Yesterday’s blog investigated the way our brains might be susceptible, due to cognitive biases, to bad or disastrous news. But sometimes, even if you can master your reactiveness, you can still see how one visualisation of a set of facts can be more useful to the powers that be than others.

For example, look at these below:

Red is Asia/Pacific; Blue is Africa; Green is North/South America; Yellow is Europe. Incomes are PPP.

Red is Asia/Pacific; Blue is Africa; Green is North/South America; Yellow is Europe. Incomes are PPP.

As Jason writes in this blog, this has been called the “two-hump world” - where it looks like the developing world has seriously caught up with the developed world, on income inequality. In terms of the growing power of “Asia/Pacific” (the red zone), who could possibly deny the claims from gurus that we can’t talk about North-South poverty anymore? Hickel shows the chorus that’s used this image:

In his post on inequality, Roser uses this graph to conclude: “The poorer countries have caught up, and world income inequality has declined.” Hans Rosling went further, saying that thinking about the world in terms of North and South is no longer a useful lens, as the South has caught up to the North. Bill Gates has used the graph to claim that “the world is no longer separated between the West and the Rest.” Steven Pinker leveraged it for the same purpose in his book Enlightenment Now. And Duncan Green recently wrote that income inequality is no longer about a divide between nations or regions of the world, but rather between social groups within the global population as a whole.

The problem is, says Hickel, is that the reality - according to another view of poverty stats - is exactly the other way around. Global inequality has been rapidly increasing over this period.

The above graphs are wrong in three ways, says Hickel:

  1. The vertical axis compresses the world’s populations into a hump, defined by a ratio (logarithmically). Do it in a linear way, and the populations of the world spilt into each side of the table - a huge bulk at the poorer end, and a long ascending tail, shooting upwards after 30 dollars a day.

  2. The method they use for comparing prices across all these decades (something called Purchasing Price Parity) seriously over-estimates the spending power of the poor.

  3. And if the point is to compare Global North and Global South, then their regional groupings are all wrong. Really, if you want to use “Asia and Pacific” as a category to show the income growth in poorer parts of the world, should you include Australia, New Zealand and Japan in that? How useful is it, if you want to measuring the change of income between richer and poorer regions, to include Haiti and Belize, alongside the US and Canada, in “The Americas”?

Fix all these issues - make a clear divide between Global South and Global North; don’t overestimate spending power of the poor; and put the populations that relate to these incomes on a straight line, rather than a ratio to each other… And you get the animation below:

As the populations grow, the bubbles grow. But note the grey bubble (the Global North) getting bigger, and streaking off to the right, as income inequality (based on income per head of population) increases and increases. Hickel explains:

The income gap between the average person in the North and the average person in the South has nearly quadrupled in size, going from $9,000 in 1960 to $35,000 today.

In other words, there has been no “catch up”, no “convergence”. On the contrary, what’s happening is divergence, big time.

This is not to say that Rosling and Roser’s hump graphs are wrong. They tell us important things about how world demographics have changed. But they certainly cannot be used to conclude that poor countries have “caught up”, or that the North-South divide no longer exists, or that income inequality between nations doesn’t matter anymore. Indeed, quite the opposite is true.

You can go to the post itself to see Hickel’s explanation for why this inequality has not actually lessened, but increased. But we present this more as a fascinating case study in the politics of visual knowledge - and we would invite comment as to whether Hickel has himself smuggled in contestable assumptions. “Slow truth” is what we prefer.