Sources of racial disparity in arrest and incarceration: a critique of Beaver et al. (2013)

The paper by Beaver et al. (2013) makes multiple false assumptions to reach an inaccurate conclusion, which is at odds with multiple previous studies on the subject of racial disproportionality in the criminal justice system. The first of these assumptions is that all crimes can reasonably be lumped together in the category of "violence", and that all of them are equally likely to result in arrest and incarceration. The second is that violence is a reasonable way to approximate the extent to which one engages in criminal activities, as well as the severity of these activities.

The authors base their analysis on the assumption that individuals with lower IQs are more likely to be arrested and incarcerated, as are individuals who commit more serious (i.e. severe) crimes. As they state on page 30 of their paper, "...both offense severity/frequency and IQ correlate with criminal justice processing, with...individuals with lower levels of overall intelligence being more likely to be arrested and processed by the criminal justice system." The second of these assumptions is hardly debatable, and is clearly supported by the studies the authors cite (Beaver et al., p. 30). However, the assumption that individuals with lower IQs are at a higher risk of arrest and incarceration is not supported by the studies cited by Beaver et al. For example, the first one they cite in support of this claim is Diamond et al. (2012), which shows that prison inmates with lower IQs commit more violent acts in prison. This, however, is not the same as being more likely to be arrested or incarcerated, partly because committing violent behavior in prison, obviously, cannot lead to either outcome if the perpetrator is already incarcerated. Similarly, the authors misrepresent the results of another study they cite for this claim, which showed that "Low IQ was related to antisocial behavior at age 5 years and predicted relatively higher antisocial behavior scores at age 7 years when antisocial behavior at age 5 years was controlled" (Koenen et al. 2006). This is, again, hardly the same as being more likely to be processed through the criminal justice system. These misrepresentations undermine the authors' claim to be controlling for factors (other than race) that can influence the likelihood of such processing.

The authors also assume, absurdly, that their scale measuring lifetime violence is an effective way to approximate the frequency and severity of an individual's criminal offending. Previous studies examining whether differences in offending explain racial differences in incarceration rates have used a more detailed and accurate methodology that does not lump all criminal offenses into the category of "violence". For example, Langan et al. (1985) estimated the number of blacks expected to be incarcerated by "multiplying the number of black offenders by the crime-specific probability of a white offender going to prison" (p. 678). He then compared these data to the percent of prisoners serving time for a specific offense to see if blacks were overrepresented to the expected % of black prisoners for that offense (Langan et al. 1985, table 4, p. 677). Langan found that "even if racism exists, it might explain only a small part of the gap between the 11% black representation in the United States adult population and the now nearly 50% black representation among persons entering state prisons each year in the United States" (p. 682). Lastly, Beaver et al. provide no source for their estimate of the magnitude of the violence-arrest/incarceration relationship, nor for the IQ-arrest/incarceration relationship, which is essential to know that they are controlling for these factors correctly.

These methodological issues undermine the conclusions of Beaver et al.'s paper.

Sources:
Beaver et al. 2013: (abstract, full text download link)
Diamond et al. 2012: (abstract)
Koenen et al. 2006: (PMC full text)
Langan et al. 1985: (PDF)