No, that isn't true at all. Employers can only hire from the pool that exists. If the applicant pool is 90% white dudes, then an unbiased hiring process will result in a very skewed employee demographic. Assuming bias based on differences in statistics is motivated reasoning and should not be entertained.

And why would the applicant pool be 90% white dudes?

Because the non-whites and non-Asians can't pass academic tests as easily?

And then why are non-white and non-Asian students not passing academic tests at the same rate as their counterparts?

Nobody really knows. A mixture of their environment, parenting, and genetics probably. I don't know what extent of which and I don't think anyone does.

> Nobody really knows

That's not quite true.

> A mixture of their environment, parenting

One would wonder what could be behind that... Also, poorer schools in poor neighborhoods because they receive less funding, lowering the value or real estate in the neighborhood, attracting poor people to it.

> and genetics probably

Funny how this always comes up.

Oh wow, if nobody really knows that sounds like a big problem we should solve! Maybe we should fund research into that. And maybe fund programs to alleviate the environments that were created. I'm sure anyone who wants America to be a true meritocracy wouldn't be rooting out this big unsolved problem.

If some racial groups on average are a standard deviation below 100 IQ, then they would tend to have lower representation on merit basis.

There are some positions that you don't want low IQ (low intelligence) people doing them.

If a group of 'X' people (based on any factor you like, be it race, gender, ideology, protected characteristics, etc) are statistically lower IQ, you would expect this to be reflected in selection over a large enough sample.

Equally, I would expect that when you're fighting for people who are in high demand, that the wage reflects it. Merit is a "positive discrimination" for people who do the job better.

> Considering color or religion doesn't have any impact on the ability of someone to be a "best astronaut", if your groups isn't diverse, it's most likely there is a bias in action.

You have 8 blue people and 2 purple people. The blue people and purple people have different colour, different ideology, but are the same in all ways that count.

The astronaut programme opens up and needs 5 people. From 10 people, choose 5. It turns out that there are 252 ways of doing this [1]. At most you could only have 2 purple people selected, when 1 purple person is selected it looks unbiased, but actually the vast majority of cases are when only blue people are represented.

Another way to think about it is this: 8 blue people, 2 purple people. You pick one at random, there is an 8/10 chance you pick blue, 2/10 purple. If you select a purple, for the next pick there is now just a 1/9 chance of picking purple, but an 8/9 chance of picking blue.

[1] https://www.hackmath.net/en/calculator/n-choose-k?n=10&k=5&o...