Does inequality cause crime?

Thursday, May 4th, 2017

It is often argued that inequality or poverty causes crime, Devin Helton notes, and that only by addressing these “root causes” can we reduce crime:

But as we know from Statistics 101, correlation does not prove causation. There are numerous variables that differ between countries or states that can have a correlated impact on inequality and crime: governance, ethnicity, culture, institutions, traditions, etc.

To avoid these confounders, another way to test the link between inequality and crime is to examine the treatment effect. If public policy choices increase inequality, does crime go up? If public policy choices decrease inequality, does crime go down?

From 1910 until the late 1970s, both England and America undertook concerted programs to reduce inequality. Both introduced progressive income taxes. Both changed laws to support unionization. And then in the 1980s both countries reversed course. Britain elected Thatcher, the U.S. elected Reagan. They lowered tax rates, made life more difficult for unions, and promoted business. Inequality rose in both countries for the next few decades.

I hope you see where this is going:

Turns out that inequality reduces crime, and equality increases crime. For every 10% decline in inequality according GINI coefficient, homicide nearly doubles! That is a very strong correlation. (You can download my spreadsheet here).

Does he actually believe this?

No. My results above are due to tricks and confounding factors.

[...]

In total, the statisical analysis above does not prove causation. But — all those studies using correlations to show the opposite, that inequality causes crime, are also bogus. They are also cherry-picked, confounded, and intellectually dishonest. With so many interlocking causal factors, anyone who calculates a correlation with regards to inequality and crime and tells you this proves X causes Y is either appallingly stupid or utterly mendacious.

Read the whole thing.

Leave a Reply