Friday, February 17, 2012

Gini Coefficient Games (Or why John Rentoul is wrong on inequality)

John Rentoul is under the impression that the last 30 years haven't seen massive rises in inequality and cites an ONS report detailing the changes in the Gini coefficient for the UK that shows inequality hasn't really moved since the early 90s. Not a point I particularly agree with, also I find it a little saddening that centre left types like Rob Marchant seem to be cheering him on (surely it's the Tories, not the combined forces of Owen Jones and Compass we should fighting).

To demonstrate my point I constructed a quick Gini coefficient calculator and constructed a set of income data with the total income divided in the following proportions:

Top 0.1% - 7%
99-99.9% - 8%
90-99% - 21%
75-90% - 16%
50-75% - 21%
25-50% - 16%
Bottom 25% - 11%

The Gini coefficeint for this society is 0.369, it's more unequal than the UK, but not a million miles off. We can now play a bit of a what if game, so, first off lets shave 1% of each of the three bands in the 50-99% range and give it to the top 0.1%, this gives us the income distribution below:


Top 0.1% - 10%
99-99.9% - 8%
90-99% - 20%
75-90% - 15%
50-75% - 20%
25-50% - 16%
Bottom 25% - 11%

The effect of this to raise the Gini to 0.382. Now, let's take this initial change a little further, let's try shaving 1% from the 25-50%, 50-75% and 75-90% bands and and adding them to the bottom 25%. The numbers we get as a result of this change are shown below:

Top 0.1% - 10%
99-99.9% - 8%
90-99% - 20%
75-90% - 14%
50-75% - 19%
25-50% - 15%
Bottom 25% - 14%

The Gini coefficient of this set of income distributions is 0.35, by squeezing those in the middle income bracket we have managed to lower the overall measure of inequality. The point of all this is to demonstrate that the Gini coefficient has it's limits as a measure of inequality. You can maintain a given Gini coefficient if you counter a rise in the incomes of the top 0.1% or 1% by narrowing the gap between the poor and the middle classes.

Now maybe all this is a roundabout way of getting to John's final point where he admits that the incomes of the top 0.1% have risen massively. But it is somewhat misleading in the rest of his post to suggest that the wealth of work on inequality has been concerned with Gini coefficients rather than the income of the top 0.1%. As to his final dismissal, I would ask why he believes that it is necessary to accept these vast rewards given that our pre 1980 economic growth record (when these rewards were a rarity) was actually a good deal better.

1 comment:

Unknown said...

Sorry if my blog post was not clear, but I did not say that inequality has been unchanged for 30 years. Quite the contrary. It increased in the 1980s and since about 1990 has been broadly unchanged.