Is Twitter more unequal than life, sex or happiness?
Bill Gates has more money than I do. JK Rowling sells more books. Katy Perry has more Twitter followers. Usain Bolt is faster. And I can only presume that Buddhist monk Matthieu Ricard, reportedly “the happiest man alive”, is more cheerful.
The world is an unequal place. Exactly how unequal, though, depends on what we measure and how we measure it.
Researchers concerned about the concentration of money in the hands of a small number of people tend to focus on the income or wealth share of high earners. In the US, the income share of the top 1 per cent has soared from 11 per cent in 1980 to 20 per cent in 2016, according to the World Inequality Report 2018. But in western Europe it has merely moved from 10 per cent to 12 per cent.
The top income share highlights something important, but it misses changes elsewhere in the income distribution. If you want a single number to summarise the whole distribution, the natural choice is the Gini coefficient, which varies between zero (equality) and 100 per cent (one person has everything).
The coefficient of post-tax income in the US is nearly 40 per cent, and has been increasing for years. This rise — alongside the increase in the US top income share — has created a perception that inequality is rising everywhere and by every measure. That is not true.
Globally, the Gini coefficient of income is a shockingly high 65 per cent, but it is falling, helped by strong growth in large emerging economies. In France, the Gini coefficient of income has been broadly falling since the 1950s; it is now about 29 per cent.
In the UK, the Gini increased sharply in the 1980s, but not since; it is around 35 per cent today. And as the website Our World In Data observes, the Gini coefficient was also much higher a couple of centuries ago in the UK, perhaps 50 or even 60 per cent.
So that is the story for income inequality. But the Gini coefficient can be applied to inequality in any set of numbers you like, from the number of storks in each country to the body weights of a family of hippos. For example: authoritative data on sexual activity in the UK are available from Natsal-3, the third National Survey of Sexual Attitudes and Lifestyles. Natsal-3 reports the number of opposite-sex partners we say we’ve had in our lives, and the number of times we say we’ve had heterosexual sex in the past four weeks. (It will surprise nobody to hear that men and women make rather different claims, so I’ve averaged their responses.)
Since I know you may be curious, I have made my own calculations, based on these data. For 35-44 year olds, the Gini coefficient of recent sexual activity is 58 per cent. The Gini coefficient of lifetime opposite-sex partners is lower: 50 per cent. Both are much higher than income inequality in the UK.
Nor are these figures driven by a few outliers with thousands of partners. When it comes to the bedroom, we don’t need to consider extremes to witness considerable inequality: many perfectly ordinary people have had only one sexual partner, or none, and many perfectly ordinary people have had at least 10. Bigger variations exist in income, but only at the extremes of distribution.
Of course, while one can measure income and sex using the same statistical method, that does not mean the moral or political implications are comparable. Most of us wouldn’t mind having more money, but it is far from obvious that we all want more lovers. Who has the time?
I would be pleased if the Financial Times decided to increase my salary, but unsettled if they set up a profile for me on Ashley Madison, the marital affairs site. And while the government can redistribute money, it cannot redistribute consenting adults.
There is more to life than money and sex. What about — well, life itself? Some people die young; others endure.
Researchers have computed Gini coefficients for longevity. The economist Sam Peltzman found that inequality of life expectancy has declined enormously since the mid-19th century. The Gini coefficient was then around 50 per cent in the US, because so many people died in infancy. It is now about 10 per cent.
Across the world — according to Jeroen Smits and Christiaan Monden — longevity inequality is just 18 per cent. Looking only at those who survive to adulthood, it is 12 per cent. Recall that the Gini coefficient of income is 65 per cent globally. Since life cannot be reallocated, it is cheering to see that it is far more equally distributed than income.
I have given into temptation in comparing Gini coefficients across domains. Such comparisons can lead us astray. Yes, Perry has a thousand times as many Twitter followers as I do, but Bolt will never be a thousand times faster, nor Ricard a thousand times happier — not according to the oft-used life satisfaction scale of 0-10, anyway. Does that make Twitter more unequal than happiness? I am not sure.
There lies a statistical lesson: measuring inequality is just a step on the road to understanding it. Or so a Zen statistician might say.
Written for and first published in the Financial Times on 17 August 2018.