Open rickecon opened 4 years ago
@MaxGhenis Add to these the Lorenz curve, Gini coefficient, wealth shares, and variance of log wealth. These are common measures of inequality. And I like the way that OG-USA puts these in an Inequality
class object.
wealth.py
module, lines 133-137 and in the Inequality
class of the utils.py
module, lines 469-501. The Wikipedia Gini coefficient page has some nice expressions for the equation for computing the Gini coefficient.x
% of the wealth is held by the first x
% of the individuals for all x
). A right angle from the lower left origin to the lower right origin to the upper right corner signifies ultimately concentrated wealth (i.e., 0% of the wealth is held by the first x
% of individuals for all x
except for the wealthiest individual who holds 100% of the wealth). The Gini coefficient above is defined as (area A) / (area A + area B). Zhiya Zuo has a blog post on how to plot the Lorenz curve using Python.
y
th percentile bin, where y
is defined as the x_j
th percentile minus the x_i
th percentile. The deciles, quintiles, and quartiles are a good examples. OG-USA uses the following quantiles [0.25, 0.25, 0.20, 0.10, 0.10, 0.09, 0.01]
.
@MaxGhenis I have been compiling a list of figures that I would like to be able to quickly make. Here are all the ones I have collected so far.