The Justice System is Explicitly Racist
It's been telling us for years.
You can download the data and code I used to make the analyses on my GitHub.
TL:DR - Data from the General Social Survey shows that people working in the justice system tend to be more conservative than other workers, and they also endorse more racist views against blacks compared to whites (30% vs. 16%).
Compared to other workers they were also less likely to blame the poorer outcomes experienced by African-Americans on discrimination, although they were no more likely to blame it on lower innate intelligence, less opportunity for education, or lack of will.
Justice workers were no more likely to express negative views towards whites or blacks compared to other workers, and were less likely to express negative views towards Hispanics, although there was less data available for this group. They were also less likely to express positive views towards blacks (33.6%) and Hispanics (25.3%), both compared to whites (43.8%), and to other workers.
Justice workers were also more likely to say that immigration should be reduced rather than increased compared to other workers.
There is little agreement on the prevalence and pervasiveness of racial bias in the justice system. While some point to the fact that police kill black Americans at nearly three times the rate of white Americans as clear evidence of racism, some scholars believe that this is due to differences in crime rates between white and black Americans, while others argue that this conclusion is a result of improper statistical analysis.
You might’ve heard some people say that non-white defendants are given longer sentences compared to whites, although others argue that this disparity may be because of other factors, like the socioeconomic background of the defendant or type of weapon used.
Similarly, there is little agreement on the extent to which the race of the perpetrator and the victim affect clearance and conviction rates of homicides.
We can talk about what might be the best methods to measure these things, flaws in the data, what control variables to use, but arguably the simplest way to assess racial bias in the justice system is to just ask the workers in this field.
The General Social Survey (GSS) is a survey of U.S. citizens which asks thousands of participants questions about themselves and their views on society, and these include questions about their views on race and in what area they work - and some of the participants work in the justice system. Since the GSS is meant to be a random and representative survey of Americans we can assume with some degree of certainty that the justice workers interviewed will also be fairly representative of their field. This gives us the opportunity to assess the racial biases of workers in the justice system and compare them to other workers.
The GSS uses the 2007 North American Industry Classification System (NAICS), to group the workers into specific areas. The group for workers in the justice system is “Justice, Public Order, and Safety Activities”, which includes the following areas: Courts; Police Protection; Legal Counsel and Prosecution; Correctional Institutions; Parole Offices and Probation Offices; Fire Protection; and Other Justice, Public Order, and Safety Activities. You can see more details in the NAICS website linked above.
The questions that I’ll include in this analysis asked participants for their views on the following topics:
Political views on a liberal-conservative scale;
If the participant believes immigration should increase, reduce, or stay the same;
Whether they ascribes the worse outcomes black Americans experience to discrimination, lower innate intelligence, less opportunity for education, or lack of motivation;
Where the participant rates whites, blacks and Hispanics on a hard-working vs. lazy scale;
Where the participant rates whites, blacks and Hispanics on an intelligent vs. unintelligent scale;
Whether they would oppose a close relative marrying a white, black or Hispanic person.
At the end I’ll also show how many of the respondents endorsed at least one negative and one positive view against each of the racial groups by considering points 4-6, as point 3 only asks about African-Americans. Let’s get on to the graphs.
First I was curious to see how the political views of justice workers compares to that of other workers. As you can see in the graph below, they tend to lean more towards conservatism.
When it comes to immigration, justice workers were more likely to say immigration should be reduced rather than increased, and were more likely to say so compared to other workers.
Next are the questions pertaining to point 1 - why the participant thinks African-Americans tend to have worse outcomes. We can see that justice workers are less likely to blame these differences on discrimination, however they were also no more likely to ascribe it to lower innate intelligence, less opportunity for education, or lack of will.
Next come the questions on how hard-working or lazy the participant believes certain groups of people to be. Compared to other workers, justice workers were more likely to rate whites neutrally, and were less likely to rate them as lazy. They were also more likely to rate blacks neutrally, but were less likely to rate them as hard-working. Their answers regarding Hispanic Americans were similar to other workers’, but as these questions were only asked from 2018 onward there is less data on them.
Next come the questions on intelligence. Justice workers were more likely to rate whites and Hispanics neutrally, but for blacks their answers were similar to that of other workers.
Next come the questions asking if the participant would support or oppose a close relative marrying a person of the different racial groups. Overall the answers of justice workers were similar to other workers, save for them being less likely to oppose a relative marrying a Hispanic person.
Finally, I aggregated all the answers to see how many respondents report one or more negative views towards the different racial groups, and how many report one or more positive views towards these groups. I considered negative views rating a group as lazy, unintelligent, or opposing that a close relative marry them. Positive views were the opposite: Rating a group as hard-working, intelligent or favoring that a close relative marry them.
Overall, justice workers were more likely to have negative views towards blacks compared to whites (30% vs. 16%), and were also less likely to report negative views towards Hispanics (7.4%), both compared to other racial groups and to other workers. They were also less likely to report positive views towards blacks and Hispanics compared to whites, and compared to other workers.
Because some of the questions about Hispanics were only asked from 2018 onward the sample of justice workers was smaller, so the generalizability of the results regarding attitudes towards Hispanics is more questionable.
Though some of the questions on Hispanics were not asked prior to 2018, the GSS had been asking about blacks and whites since the 90s, so I also plotted the results of these questions for the period between 2000-2012, which you can see in this Imgur album. Overall the pattern of their answers were fairly similar.
Now, it’s worth pointing out that these questions will obviously not be a perfect measure of racist or xenophobic attitudes. I included the variables on the topics of intelligence, work and marriage because the GSS asked them about the three racial groups I analyzed, so I could compare them to each other. There were other items and racial groups that I couldn’t include in this analysis because the questions were either asked about only one group or were asked in different years than the ones I included. For example, if we also include those who believe that blacks have lower innate intelligence in the total (‘racdif2’ variable), the number of racist justice workers increases slightly to 32%.
A thorough exploration of racism and xenophobia would require a lot more questions - but we would also need to agree on what questions should be asked first. There are beliefs that most of us will agree are explicitly racist, but there are things that are implicitly racist without most people realizing that is the case.
One example of this is the variable ‘racdif4’. Putting aside the question of whether African-Americans actually have less willpower or not, most people who say that black Americans just need to have more motivation, that they should “pull themselves up by the bootstraps”, would likely not describe themselves as racist or explicitly say that blacks are inferior to whites. But who we are, including our willpower, is determined by nature (genes) and nurture (environment), so in insisting that the worse outcomes experienced by blacks are not a result of a discriminatory environment they are inadvertently making the argument that this supposed lower willpower is a result of nature, that African-Americans are just inherently less motivated.
Next we have the problem that, because racism is socially undesirable, some people simply don’t answer these questions honestly. Several studies have found that when using indirect questioning techniques participants are more likely to endorse racist, antisemitic and xenophobic attitudes. A study by Pew found that 48% of whites, 29% of blacks and 42% of black-white biracial people had some degree of subconscious bias against blacks - yet they found little to no relationship between the results of the Implicit Association Test and explicit questions about racial prejudice. There are legitimate criticisms to be made of the IAT, but as the Pew article points out, several studies have found that the IAT has good predictive validity, and sometimes even exceeds the validity of explicit measures of racism.
Measuring racism is more complicated than it might seem, which is why, as mentioned earlier, while most researchers agree that there exists racial bias in the justice system, there is still plenty of disagreement. Neither this analysis nor any other will be the ultimate, perfect assessment of racism in the justice system. But it’s important to explore this topic through various means to understand it more thoroughly and add more evidence to the discussion - and much like other research, results from the GSS suggest that racism is still very much prevalent in the justice system.