> if the population of people you hire is out of proportion with the general population.

why it should be in proportion? This argument is flawed as it ignores that not everyone in general population has the same abilities.

You can look at the roster of any top performing NBA team and tell me if their players population represents general population.

Companies and organizations want to hire the best for any given job, and given this preference, there will always be discrepancy btw hired and general population.

In fact the only case when both population would be equal, is if hiring decisions are based on random chance

> Companies and organizations want to hire the best for any given job, and given this preference, there will always be discrepancy btw hired and general population.

This is an assumption, not a fact. Companies hire less-qualified candidates all the time through nepotism.

I have a mixed opinion on this. There are cases when nepotism fires back via poor job performance and quickly self-corrects itself.

But there are cases when nepotism is a job requirement, as in being connected to a powerful network is job description. You can find it in investment banking, big law, VC, sales when people are hired because they know the right people and can get the right results through back-channels/established network.

You cannot make the same argument however in favor of hiring lesbians, unless being a lesbian is a job requirement

> DEI can be implemented badly, but that doesn’t mean that DEI is bad

I really doubt that DEI per your definition above can be implemented for a general benefit. Fundamentally you need to show that not hiring the best candidate is best. This is typically squared by saying that group dynamics and output improve with diverse viewpoints, which I actually agree with within limits (the limit being that if nobody agrees nothing gets done). But here we made a sleight and equated diversity based on innate properties to diversity of viewpoints, which are different things and it’s actually rather stereotypical to equate them.

I mean, you have the same problem. What does 'best' mean, and how do you accurately measure it?

Of course you will have an answer, but how various things are weighted when making a decision is pretty subjective.

Hiring will always have a subjective aspect to it, which is just the nature it. And I do agree that techniques like being conscious of bias and trying to interrupt it are useful.

But yes, I do have an answer, which is that excluding factors that do not relate at all to job responsibilities is strictly better than including those, even if the result is still subjective.

Categorizing factors as not related to job responsibilities is also subjective!

Of course it can often be easy to identify the most salient factors, but that doesn't make it a concrete activity.

are you making a claim that sexual identity and race (two most common DEI factors) are part of the job ?

how is being a black lesbian part to any job whatsoever ?

No, I'm making a claim that a sloppy argument was made.

You are focusing on factors that you consider obvious, which is exactly my point, using your subjective judgement doesn't give you a framework that isn't subjective.

You have constructed and debating a strawman.

Everything involving human judgement will be subjective, yet we still can distinguish between good and bad decisions and discuss tradeoffs about being less wrong while being subjective

I don't think it's a straw man. No one has actually solved hiring, "best" isn't defined.

Like people are freaking out about any sort of DEI factor even getting considered and then actually hiring based on how Phil vibes with the candidate. Good stuff.

Can I mathematically disprove that if I take the Lagrangian of the standard model and combine it with with Einstein's field equations, and chase down emergent complexity over many levels, that the properties that DEI selects for are actually what makes a great SDE (in my world)? No I cannot. But we're getting silly here.

I will settle for the most salient factors if that sounds more precise to you. I will interpret that as a large number of sigmas, but we don't have to agree on the exact number of sigmas here!

we can go from the other side. If you want to hire the best, you need to stop looking at skin-level shallow factors and stop hiring decisions based on skin/race/sexual identity/ethnicity.

Just fair competition and pick the best candidate you think will do the job best. If you see the population is unbalanced (too many males or whites for example) => you go to the root of the problem and address the underlying root cause, not try to cover up the discrepancy by reverse discrimination at the end of the hiring pipeline.

There are a lot of societal problems in America and under-represented minorities absolutely have it worse than the generally well-off white population. But this is the argument to go and fix the root cause: income inequality, education inequality, lack of jobs, housing, etc.

The moment America decides to improve poor people's lives, not just promote and hire rich people with the "right" diversity traits, is when the whole country will improve in terms of outcomes

And a true Scotsman would know this.

Why isn’t DEI being used to hire men in female dominated industries?

It is.

Here is a supporting example found from a DDG search for "DEI nursing":

"The Rise of Diversity, Equity, and Inclusion (DEI) Practitioners in Academic Nursing" at https://ojin.nursingworld.org/table-of-contents/volume-27-20...

> Efforts to strengthen diversity in nursing have not targeted race alone. That is because the nursing profession also lacks gender diversity, with the vast majority of nurses (over 87%) being female (U.S. Bureau of Labor Statistics, n.d.). In addition, the racial demographics of faculty reflect a body of academicians who are 82% White, 9% Black/African American, 3.4% Hispanic/Latino, 2.9% Asian, and 93% female (NLN, 2021). Ideally, the nursing workforce should closely align with the racial, ethnic, and gender makeup of the U.S. population.