On Poverty and Crime

I am going to be looking at theories which state that economic factors explain part, or all, of the difference in the rates at which blacks and whites commit crime. Such theories typically concentrate on poverty and income inequality. Many proponents of such theories will argue that poverty, the state of having few resources, and income inequality, the state of having few resources relative to others in your community, cause resentment that lead to crime. And the alternative to crime, for the poor and unemployed, is less appealing than the alternatives enjoyed by the rich and those on top of the income distribution. This, combined with the fact that poverty, income inequality, ect, are all more common among Blacks than among Whites, leads to the conclusion that these factors explain part of why Blacks commit more crime than Whites do.

The first step in assessing this theory is to determine whether or not a cross sectional correlation exists between poverty and crime. In other words, is there more crime in poorer countries, counties, and states, than in richer countries, counties, and states. Given that the truth of this theory is often regarded as common sense, you would expect that this bare minimum requirement of evidence would be soundly satisfied. Surprisingly, the data is actually very mixed. One paper, published in the journal social pathology, looked at 45 different studies that assessed whether or not there was a correlation between income inequality, absolute poverty, and violent crime at state, metropolitan area, country, city, and neighborhood levels. Across these 45 studies 68 total correlations were found between absolute poverty and various violent crimes (rape, robbery, murder, ect). Of these 68 only 25, or 37%, statistically significant and positive. Across these studies there were 59 correlations found between income inequality and violent crime. Of these, 27, or 46%, were statically significant and positive. This means that the majority of these studies failed to find a statically significant positive correlation between income inequality and crime as well as absolute poverty and crime. Moreover, there was a great deal of variation in the results of these studies. This is partly explained by the fact that there is no wide agreement between researchers on what level of aggregation to look at when measuring these variables. Some researchers, for instance, looked at states while others looked at counties. Still, this wide variation in results suggests that when a strong positive relationship is present it is likely explained by some other confounding factor. After all, if poverty or income inequality really caused crime we would expect a large majority of cross sectional studies to find that areas with high levels of income inequality and poverty have high levels of crime. On the other hand, if there was some other variable which caused both poverty and crime then we would expect to only find such a correlation when that third variable was present. This could explain the variability in results.

Also of interest here is the fact that the cross sectional relationship isn’t consistent across the races. For instance, a 2002 paper published in “Population and Environment” found that a negative relationship between wealth and the overall crime rates in White and Asian populations at the country level but that there was actually a positive relationship in African and Caribbean countries. This means that the richer a Black nation was the higher its crime rate was! To conduct a similar analysis with more up to date data, one can take the Average homicide rate estimate in black countries from Interpol as-well as the GDP per Capita figures for those nations from the World Bank, and one will find that no statically significant relationship exists between these two variables. This does not suggests that the relationship found in previously referenced paper no longer exists since homicide rates aren’t the same as general crime rates, but it does cast further doubt on the idea that poor economic conditions cause blacks to commit more crime.

Another way to test this theory would be to look at longitudinal data. One important prediction of the theory that poor economic conditions cause crime is that crime will rise when the economy gets tough and lower then it runs smoothly. The recent history of America attests to the fact that this isn’t a relationship that you can count on. During the great depression, the worst economic time the US’s history, crime rates fell. During the 1960’s, arguably the best economic time in the US’s history, crime rates rose. During the 2008 recession, in which their were record levels of income inequality and a 30 year low in labor participation, we saw the lowest violent crime rates of the last 40 years. And a recent paper which looked at the relationship between income inequality and violent crime rates in the US from the year 1981 to the year 1999 found a strong negative relationship between the two. Income inequality rose sharply over that period while violent crime plummeted. It is true that some studies, especially ones looking at very small levels of aggregation, have found results that support this theory. The evidence at that level is very mixed. But the extent to which this theory does not match up with national data suggests that it is not an overwhelming determinant of crime. And the variability of the results at lower levels of aggregation suggests the same problems that the variability of cross sectional data suggested.

Thus, this theory doesn’t even get off of the ground. In general, it isn’t clear that poorer regions have more crime than richer regions, and it especially isn’t clear that poorer black regions have more crime than richer black regions. Nor is it clear that crime consistently rises when times get tough. But even if it was the direction of causality would still be unclear. That is, would poor economic conditions be causing crime or would crime be causing poor economic conditions? Either direction is intuitively plausible. It could be that being poor makes people resentful and, seeing that they don’t have much in the way of other opportunities, they turn to a life of crime. But it could just as easily be the case that crime causes businesses to leave a community, resulting in unemployment and that those who resort to crime no longer report a large amount of their income to the government, resulting in increased poverty and income inequality statistics. The list of possible explanations either way could go on forever.

It is important to note that if you are a proponent of this theory, in spite of a complete lack of evidence, this theory can only work as a partial explanation of the racial crime disparity. It cannot be a complete explanation because the racial composition of a region is actually a better predictor of its crime rate than poverty, income inequality, education levels, ect. In other words, when it comes to a places crime rate: race matters more than economics.

To sum up: the only thing that this theory has going for it is that it is taken for granted by an enormous number of people. Poorer areas are not consistently more crime ridden than less poor areas, and rich black nations are sometimes more crime ridden, on average, than poor black nations. Economic down turns do not consistently coincide with crime waves, nor do economic booms consistently coincide with down turns in crime. And even if all this imaginary data did exist the direction of causality would remain unknown. There doesn’t seem to be any good reason to believe that economics plays an important role in racial crime disparities.


  1. Raw data for correlation between Avg. Homicide rate and GDP per capita:

  2. Analysis of 45 studies on the relationship between income inequality, poverty, and violent crime:

  3. The positive association between general crime rates and wealth in Black nations:

  4. Crime rates and economic conditions in recent US history: