Welcome to our community

Be a part of something great, join today!

Discrimination and Disparities, by Thomas Sowell

  • Thread starter
  • Admin
  • #1


Indicium Physicus
Staff member
Jan 26, 2012
I just finished reading Discrimination and Disparities, by Thomas Sowell. Whatta book! I would highly recommend it. The main point of the book is to show that disparities do NOT imply discrimination (oppression). To show this, Sowell runs through a huge number of examples, both in nature and in human relations, where disparities have existed and discrimination is demonstrably absent.

To be sure, disparities are sometimes caused by discrimination, and Sowell mentions that: disparities do not imply a LACK of discrimination, either. However, in today's culture, that is not the main fallacy being committed. I see the disparities-implying-discrimination fallacy ALL the time. For example, Sen. Cory Booker (D-NJ) questioning the Hon. Amy Coney Barrett: essentially his entire set of questions committed this fallacy.

What does cause disparities? Sowell's answer: causes can be many and varied, and are not always easy to point out. Sometimes you can point to a cause, sometimes not. But one thing you CANNOT do is simply infer discrimination or oppression as the cause of disparities.

So the next time someone tries this line on you: "People group A and people group B have such a gigantic difference in their outcomes; we have to change how we're handling these people groups in order to reduce these differences.", you need to suspect the implied link.


Staff member
Jan 26, 2012
I see this as both very true and also a way to turn a nuanced, complex set of topics into a black and white framing. Oppression is redefined constantly by the person using it and can't even be clearly defined. The murky nature of human experience is on both political sides, so if under some definition of oppression it exists at $T_0$ and doesn't at $T_n$, at what point does the oppression stop?

This idea is something you've noted strongly in the other thread too and is trivially true to me. Correlation is not causation and if event B occurs after event A, this does not show A -> B. On the other hand if there are disparities in some metric between groups and it's an order of magnitude gap, like 1% vs 10%, it is hard to claim that this is independent and clearly not stemming from history or context.

The push back on turning all things into oppression and disagreement there as part of it are both issues that are truly happening in the US where I can see it and hear about it firsthand. The other side of this same view is that it also doesn't make all oppression a problem of the past or not real. These are the two explicit objections I have with this idea and framing:
  1. Exposing an incorrect logical argument or bad point for the opposing side doesn't do anything to strengthen one's own side. It also doesn't make a range of beliefs, views, etc. collapse into two binary poles.
  2. Do you think that this argument changes anyone's mind? Is the goal to be right or to help show the flaw that you see in others' views?