January 25, 2017

Race, Poverty, and Crime

Summary: Many people believe that racial disparities in crime can be explained by the fact that Black people are, on average, poorer than White people. However, as reviewed in another article on this site, there isn’t strong evidence that poverty causes crime in general. In this article, it will be shown that race is a predictor of crime, a better predict of crime than poverty and related economic variables are, and that race continues to predict crime even when the effects of poverty are controlled for. 

Race as a Predictor of Crime

At the individual level, the relationship between race and crime is extremely well established. Ellis, Beaver, and Wright (2009) reviewed 113 studies which looked at whether or not Blacks commit more crime than whites and found that all 113 did. Similarly, all 17 studies looking at crime differences between East Asians and Whites found that East Asians commit less crime than Whites.

Ellis, Beaver, and Wright) also show that studies using self-reported criminal activity are less consistent. Some would view this as evidence that Blacks are arrested more than Whites despite committing the same amount of crime. The page “Is the Criminal Justice System Racist?”, on this site, refutes that claim.

The relationship between race and crime at the regional level is also well established. Pratt and Cullen (2005) meta-analyzed 162 studies which looked at whether regions with a greater proportion of Black people had higher crime rates. The average study found a positive effect of .294 and 72% of said studies findings were statistically significant. Out of the 34 variables commonly thought to be associated with crime that Pratt and Cullen 2005 meta analyzed, only 5, or 15%, had a greater proportion of significant findings. Moreover, none of these 5 variables had even 15 studies done on them. This likely explains their highly consistent findings. Thus, of all the heavily studied variables in Pratt and Cullen, “percent black” is the one most consistently associated with crime.

Race vs Economics as Predictors of Crime

So, race predicts crime. How does its predictive power compare to that of economic variables? Well:

Land, McCall, and Cohen 1990 collected data on the homicide rates of cities, standard metropolitan statistical areas (SMSAs), and states for the years 1960, 1970, and 1980. In each year they included all 50 states and every city and SMSA included in the census. They then looked at how well the following 11 variables predicted crime variation between these areas: population size, population density, percent black, percentage aged between 15 and 29, percent divorced, percent of kids without two parents, median family income, the poverty rate, income inequality, the unemployment rate, and whether or not the city/SMSA/State was in the south. All of these variables were entered into a single regression model, meaning that the estimated effect size for each variable held all other 10 variables constant. This analysis thus produced 9 total models explaining crime variation in cities, SMSAs, and states, across 3 decades. Across these 9 models, race was a better predictor of homicide than unemployment, poverty, and median income, in 7, or 78%, cases, and a better predictor than income inequality in 8, or 89%, cases. Thus, over 3 decades of very large data sets, race was pretty consistently a better predictor of homicide rates than economic variables were.

Another relevant analysis was carried out by Unz.com owner Ron Unz. Unz 2013 looked at how well median income, population density, poverty, and % black, correlated with the crime rates of large American cities between 2006 and 2011. He found that the size of the black population of a substantially better predictor than any of the other variables tested.

Similarly, the New Century Foundation’s report “The Color of Crime” analyzed the violent crime rates of the 50 U.S. states and D.C for the year 2005. The analysis found that state violent crime rates correlated at .81 with the percentage of the population that was Black or Hispanic, .37 with the state’s percentage of high-school drop outs, .36 with the states poverty rate, and .35 with the state’s unemployment rate.

Templer and Rushton 2011 significantly replicated the New Century Foundation’s results. They analyzed crime variation across the 50 U.S. states and found that the percent of the population that was black was a stronger correlate than average income for murder rates (.84 v -.40), robbery rates (.77 v .06) and assault rates (.54 vs -.23) The paper did find that income was a stronger predictor than black population size for rape rates (-.16 v -.22), but neither of these correlates were statistically significant or large.

Kposowa, Breault, and Harrison 1995 analyzed crime variation across 2,078 U.S counties. As can be seen below (standardized beta coefficients are under the “beta” column), the percent of the population that is black was a stronger explanatory variable than poverty, income inequality (gini), and unemployment, for explaining variation in both property and violent crime.



Finally, we have Rushton and Templer 2009 which looked at national variation in crime. They found that skin color, a proxy for race, was more strongly correlated than national income with homicide (.25 vs .17), rape (.24 vs .10), and serious assault (.20 vs .09).

Thus, across a large range a literature analyzing crime variations across cities, counties, states, and nations, we see that race is a better predictor of crime than economic variables are.

American Blacks vs European Whites

Until now, we’ve just been looking at the relative predictive power of race and economic variables. At this point, I’d like to move on from that and look at how well racial wealth disparities predict racial crime disparities.

If poverty really was the root cause of racial crime disparities, then we would expect White populations with incomes lesser than or similar to African Americans to have higher or similar crime rates to African Americans. On a global scale, this comparison is easy to make. According to the Census, the average African American makes 20,458$ a year.  According to the International Monetary Fund, the average income of a European is 25,434$. Thus, Europeans are just slightly richer than African Americans. Given this, if poverty is the main cause of racial crime disparities we should expect Europe to have similar crime rates to African Americans.

This does not pan out. The Black homicide rate in America is 34.4 per 100,000 (Cooper and Smith 2011). The homicide rate for White Americans is 4.5 per 100,000. And, according to the United Nation’s Global Study on Homicide 2013, the homicide rate for Europe is a mere 3 per 100,000. Thus, Europeans not only have lower murder rates than African Americans but also have slightly lower murder rates than White Americans. These results are hard to square with the view that Blacks have higher crime rates than Whites just because they are poor.

Racial Disparities in Crime and Economics Over Time

The changes in racial crime and economic disparities over time do not support the poverty causes black crime theory either. Over the 20th century, Black income relative to Whites has risen. Thus, on an economic theory of Black crime, we might expect Black crime rates relative to Whites to have fallen. Just the opposite has happened. Black crime rates relative to Whites have significantly increased over the 20th century and, between 1940 and 1990, have correlated at .86 with gains in Black wages relative to Whites (1).


The high value of the correlation shouldn’t be taken too seriously, as each data set only had 6 data points. But the general point remains: as Blacks have gained on Whites economically speaking their disproportional representation in crime has risen, not fallen as the economics causes black crime theory would predict.

Another Look at Regressions and a Conclusion

For the next argument of this post, we need to reconsider some data we’ve already seen. In Kposowa, Breault, and Harrison 1995 and Land, McCall, and Coen 1990 we saw that the proportion of an area that is Black is a significant predictor of crime at the county, city, SMSA, and state, level even when holding economic variables constant. Let me say that again: the blacker an area is the more criminal it will tend to be even after controlling for the effects of poverty, inequality, and unemployment.

Rich American Blacks vs Poor American Whites

Finally, we have Zaw and Darity (2016), which compared the likelihood of Blacks, Hispanics, and Whites becoming incarcerated at some point between 1985 and 2012 given what their net worth was in 1985. It is note worthy that this is the only study I am aware of which directly compares the incarceration rates of Blacks, Hispanics, and Whites, at different income levels.

Incarceration by net worth decile

Chart from Ehrenfreund (2016) data from Zaw and Darity (2016)

Clearly then, economics cannot account for all, or even most, of the racial differences we see in crime. Given its weak relationship with crime in general, and the case against that relationship being causal, it’s causal role in racial crime disparities is probably very small if it exists at all.



  1. Incarceration rates were taken from the Justice Department and the Bureau of Justice Statistics. Population size estimates were taken from the census. Racial wage disparity data taken from Chandra 2000


Facebook Comments