I last took a comprehensive look at crime seven years ago. That analysis drew on statistics for 1960-2004. The results reported in this post are based on statistics for 1960-2009. The newer analysis doesn’t contradict the older one, but it does add some explanatory power.

Cutting to the chase, the following equation explains the rate of violent and property crimes (VPC) as a function of:

BLK — the number of blacks as a decimal fraction of the population

GRO — the change in the rate of growth of real GDP per capita in the previous year, where the rate is expressed as a decimal fraction

PSQ — the square of the decimal fraction representing the proportion of the population in federal and State prisons

ORA — the number of persons of other races (not black or white) as a decimal fraction of the population.

The equation is highly significant (F = 1.44179E-31), as are the intercept and the coefficients (p-values in parentheses):

**VPC =**

**– 333768** (3.30579E-28)

**+ 339535 BLK** (1.06615E-29)

**– 6133 GRO** (0.00065)

**-174136761 PSQ** (1.00729E-15)

**– 27614 ORA** (0.0018)

(Values are rounded to the nearest whole number. The adjusted r-squared is 0.96, and the standard error of the estimate is 5.3 percent of the mean value of VPC.)

Here are the actual and estimated values of VPC:

I’m satisfied with the equation, including the negative sign on the coefficient of ORA, which mainly represents Hispanics and Asians. But I will reassess the equation if the difference between actual and estimated VPC continues to diverge after 2007, when they were almost identical.

* * *

**Related reading:** Greg Allmain, “Another Study Links Violence to the Presence of Specific Genes,” *Theden*, November 11, 2014

**Related posts:**

Lock ‘Em Up

Estimating the Rahn Curve: Or, How Government Spending Inhibits Economic Growth

Race and Reason: The Achievement Gap — Causes and Implications

Hispanics and Crime

“Conversing” about Race

Evolution and Race

“Wading” into Race, Culture, and IQ

Poverty, Crime, and Big Government