Visualizing Wealth Inequalities’ Influence on Homicides in Baltimore City, 2015 to 2019
In a previous post, I showed you different ways to visualize time data on homicides in Baltimore using data from Baltimore’s Open Data portal. For this post, I’m going to show you in one image how homicides in Baltimore are related to income, and how lower-income neighborhoods bear most of the burden of homicides in the city. First things first, let’s learn about the Lorenz Curve.
The Lorenz Curve
The Lorenz Curve is a very simple way to display the burden (or share) of something among different groups. For example, you can display the percent of total income on the Y axis and a list of countries on the X axis (ranked by a third variable of your choice), and you can see if income is distributed equally among the countries. If income is not distributed equally, then the variable by which you ranked those countries has something to do with that inequality. You will see that inequality by the curve not being a straight, 45-degree line, like this:
In the image above, income is (of course) related to income. But we could do the ranking by something else. A related measure, the GINI coefficient, can also help you quantify that inequality. This is useful if you compare two units, like Baltimore vs. Philadelphia.
Using some R programming and the Baltimore data, I created this:
On the X axis, I have ranked the 54 Community Statistical Areas (CSAs) of Baltimore (clusters of neighborhoods) by their median household income between 2015 and 2019. On the Y axis, I have the cumulative share of the 1,657 homicides that occurred in Baltimore in those 5 years.
The black line is the “line of equality.” If income had nothing to do with homicides, all CSAs would have the same share (about 1.85%) of all homicides. The red line shows that there is an unequal burden of homicides across the CSAs.
I’ve added the blue lines to show two facts: First, the less wealthy 16 CSAs had 50%…