I’ve seen a re-occurring theme in the form of a claim. The claim being that there is systematic oppression of a certain class. What people refer to as “the left” claim that minorities are systematically oppressed and therefore their realization of racism is a part of the very system they live in. Those labeled “far-right” by “the left” claim the system is working toward a white genocide where diversity and inclusion are a system to eliminate the white race. But is any of this true? Are there systems in place which aim to achieve racism or promote racists? Are there systems in place to breed out the white race? Lets examine this claim and see if a system actually exists.
The biggest flaw in this argument starts with a flawed hypothesis. In reality it’s a classic sophist utility called Composition/Division. You can read more about that here.
Allow me to walk through the most recent example given to me regarding the inflation of racist actions into a systematic racist ideological view. I’m going to form this in the frame of a study since it was a study that was presented as evidence for systematic racism.
H0: There is no systematic racism in hiring people with ethnic or minority names
H0a: There is systematic racism in hiring people with ethnic or minority names
You could expand this hypothesis beyond this of course but let’s try to keep this simple for the sake of this post. So, we must start with the Null as we would with any good study. That is the assumption of what you are testing has two possible outcomes. In this particular case you’re testing to see if systematic oppression exists or not in hiring people. Is this a good question to begin with? It seems like it’s well-formed but it’s not. There’s a glaring issue that most practiced researchers should see in the formation of this hypothesis right off the bat. In the hypothesis formation we have conflated a system with human bias and decision-making. Essentially, we’ve subsequently linked attitudes and biases of people with a system itself. You’ve successfully inserted your bias and opinion of something that exists by connecting it with human behaviors that may not reflect the system itself. It’s a very loaded hypothesis which takes two different things and attempts to merge them to prove an assumption or opinion. Here’s why.
First, the null.
There is no systematic racism… How would you test that? Would you check the people’s opinion that this exists? Or would you check the system that exists? Furthermore, who does the hiring? Is it a person or is it a system? Has the system been defined when you point to the hiring? Since people do the hiring, and people can be racist, would this then almost certainly reject the null just by racist people existing? This is how the hypothesis is malformed. You’ve connected two things, one of which we know exists, and then tested it with pretty obvious expected results to demonstrate your study rejects the null. You’ve just assured your testing will give you exactly the result you expect. Sadly, you’ve also created a worthless study that can be easily rejected because it started with a false premise in its hypothesis. However, you could have written the hypothesis like this and had a study which would give people something to think about and maybe work on adjusting behaviors rather than conflating it into a system you’re not actually testing.
H0: There is no racial bias among hiring managers when hiring people with ethnic or minority names.
H0a: There is racial bias among hiring managers when hiring people with ethnic or minority names.
Now if this were the hypothesis for the multiple resume studies that have been done, it would be a valid study. Why? Well because you have not inserted bias. The answer should be obvious at this point. You’re testing to see if there are racially biased people in charge of hiring people and if they use that bias to discriminate qualified candidates based on their name. You would be able to assume there is or isn’t racial bias there. I would love to assume there wasn’t, and wish I was right. However, the evidence from those resume “studies” rejects the null in this particular set of hypothesis. The hypothesis isn’t malformed either because it’s a very specific thing you’re testing and it’s not conflated. Say you already had the idea to do a resume study to determine if the null could be rejected or not. In this set, you would be doing exactly that. The previous had two different things being tested with a tool that would not reject or even test the null. But for my last point, this is how you would test if a system is in place that is racially motivated through a hypothesis.
Before I form this hypothesis it’s important to note the first part of the equation when forming the hypothesis. You’re wanting to know if systematic racism is real correct? How would you test that? Would you test the biases of people? Well no, people aren’t systems. People make systems that are either socially acceptable or not. We make processes and methods which can be put into a system. And sure, our biases can be put into a system. But would we test the biases of people to determine if a system has racial bias? Or do we test the system itself without the biases of ourselves or those that operate in the system? Perhaps it’s the people not following the system which create an illusion of systematic racism but the system itself doesn’t have racial bias. These are the types of questions you should ask when forming this hypothesis. So for the case of hiring people for a job, you should focus on the system people should follow when hiring people and not their biases. So what system(s) do people use to hire individuals for a job?
Well it just so happens that there’s a lot of tools out there that focus on analytics, resume parsing, word control, and more which can automate the system of finding the right person for the right job. But what happens when two identical people with identical backgrounds and education go through the system? Once more, does their names have anything to do with the system parsing resumes? Better still, what system should we test that involves the hiring process?
Well, the most common thought when it comes to hiring people through a system is best practices. Now you could test each of these for racial bias. But keep in mind, you’re going to have to test the system and not the people using the system.
“https://ideal.com/resume-screening-checklist/” (this one is in quotes because they have a WordPress bot that inserted an ad. Not necessary for this post)
For each of these processes I listed above we could test for the possibility of racial bias in the systems being presented when screening for people. In fact, we could test all of them to see if multiple systems contain racial bias thus making the decision of hiring based on race a systematic one if process X is used. Regardless, the system is the thing we’re testing, not people’s biases.
H0: Racial bias does not exist in resume screening processes or best practices
H0a: Racial bias does exist in resume screening processes or best practices
With this, we could test a system for racial biases which would reinforce racial stereotypes or even racist actions and not involve the biases of people working in those systems. So take a look at each of these systems and you tell me. Do these systems contain racial bias? Does Systematic racism exist in the hiring best practices or screening systems used by employers? It seems to me we’ve gone to great lengths to end racial inequality and destroy the concept of racism. But some people have been taught to believe there’s racism in everything and “the system” is against you because of your skin. In reality, its people that are against you. People can be racist to be sure. Hell, even systems can be racist if they’re designed to be. Segregation is an example of a system of racism. Sadly, some universities have decided, in their great stupidity, to promote segregation which is a system of race discrimination. The best option going forward is to create systems free of personal bias. We shouldn’t create systems that favor one race or gender over another. We should create systems that remove any demo-graphical information and make a decision on the merits of the person being judged. The resume “studies” are great studies if you wanted to prove that people still hold racial biases. However, they don’t test the systems in place for hiring. Also, no two people would have the exact same education and experiences. You could have the same education and job. But each experiences their responsibilities differently and would put their perception on paper. So having two identical resumes would be a control test and not a test of reality.