Comment by alaithea | Hacker News Reader

Comment by alaithea | Hacker News Reader

Parent Full thread

[edit: the author has since corrected the Gini definition in the article]

I've never encountered the Gini coefficient before, but reading up on it on Wikipedia, the author appears to have the measure flat backward.

"A score of "0" on the Gini coefficient represents complete equality, i.e., every person has the same income. A score of 1 would represent complete inequality, i.e., where one person has all the income and others have none."

Source: http://en.wikipedia.org/wiki/List_of_U.S._states_by_Gini_coe...

Also see the chart here, with the US rising toward inequality ever since the 60's. Again, the caption says "where 0 is perfect equality, and 100 is perfect inequality." http://en.wikipedia.org/wiki/Gini_coefficient#Gini_coefficie...

Glad someone else caught that. Another glaring issue with the Gini coefficient is that it tracks income inequality and not wealth inequality. It's well known that the rich don't make their money via income but rather through ownership.

TL;DR: The Gini coefficient is a shitty metric to make any conclusions, especially when taken in a vacuum.

> Another glaring issue with the Gini coefficient is that it tracks income inequality and not wealth inequality.

Well, people more frequently cite the Gini coefficient of income, but you can take a Gini coefficient of any measure, including wealth [1].

> It's well known that the rich don't make their money via income but rather through ownership.

The money made through ownership is income -- making money is equivalent to income. Its true that source of that income is different for the richest than most other people (most of the richest derive most of their income through capital, while most of the rest derive most of their income through labor), but its still income.

[1] Gini coefficients of wealth are, e.g., presented here: http://en.wikipedia.org/wiki/List_of_countries_by_distributi...

The author also claims the Gini coefficient doesn't take into count population growth and so skews the number. That is completely wrong. It's calculated using the Lorenzo curve which which is a per capita calculation.

Thanks, updated post. Conclusions were based off looking at the numbers the correct way.

Reply on news.ycombinator.com