| HN Mirror

> Right. And they're saying it would be best if they weren't accidentally doing that.

"We have already selected a male candidate for the internship so it'd be best if the next one was a woman" This is deliberately engaging in discrimination on the basis of gender. The fact that the previous hire was a man does not make it legal to use gender as a factor in hiring the next candidate, this is the gambler's fallacy [1]. I'm truly baffled as to how you think this is preventing discrimination by explicitly voicing preference for a protected class.

Your line of thinking amounts to, "we should intentially discriminate on the basis of gender, so we can be somewhat reassured we're not contributing to gender discrimination."

> Why do you think this? All efforts I've seen have been to correct for the discrimination currently occurring in the system.

Because companies often "correct" for discrimination that doesn't exist. At two of the three tech companies I've worked at, these "corrections" took the form of quotas mandating that 33% and 40% of tech hires be women, respectively. Despite the fact that ~20% of software developers and 10% of electrical engineers are women, and those two fields made up the overwhelming majority tech roles at these companies and women were already above industry average representation. This isn't correcting for discrimination. This is mandating 2-3x overrepresentation of women, this is discrimination. It'd be one thing to anonymize resumes, stripping out names, racially identifying details, and otherwise prevent recruiters and interviewers from knowing the race and gender of applicants. I'd be all for that. But that's the opposite of what DEI strives to do.

And how do you know that women are discriminated against? Do you send mock resumes to your recruiters and notice disparities between men and women? Do you anonymize interviews and notice a disparity? I doubt it. As we'll see, you reach the conclusion that women are discriminated against in a very simplistic way.

> I don't. We're still substantially less than 50% of engineering.

Correct, women make up about 20% of tech workers. Thus, a company that has 40% women in tech roles has an overrepresentation of women. Why are you comparing the demographics of a specific field with the general population? If a hospital institutes a policy that 50% of pediatricians be male, that'd be massively discriminatory against women because they make up well over 50% of pediatricians.

You seem to erroneously believe that equal employment opportunity means each job must strive to match the general population. This is incorrect. The benchmark is the demographics of the workforce, not the general population. Don't just trust me, read the law [2]:

"[Affirmative Action] is based on the premise that, absent discrimination, over time a contractor’s workforce generally will reflect the demographics of the qualified available workforce in the relevant job market."

If the workforce for a particular job is 80% women and 20% men and a company has 50% women and 50% men in that role, then that company is going to have a hard time getting Federal contracts because it's probably discriminating against women. Same if the genders were reversed. Equity with respect to the workforce not the general population is what matters.

> If this were true, men would be the minority. Women like money just as much as men do, and would leverage that non-existent favoritism to get those high paying jobs.

Disparity is not evidence of bias. Should the government mandate that women make up 50% of murder convictions? Is this just "correcting" for discrimination? No, that is discrimination. Achieving that quota would involve either convicting innocent women or deliberately letting guilty men go free, because men commit more murder than women. A non-discriminatory justice system doesn't try to achieve a predetermined outcome, it ensures that each defendant is treated equally. Quotas inevitably compromise this equality.

Women make up ~25% of the STEM workforce (software development specifically is a bit lower, around 20%). This directly matches the rate at which they graduate from STEM fields. Which matches the rate at which they say they're interested in STEM in surveys given to high school and middle school [3]. You assume that the only thing that can explain a disparity is bias or discrimination. You totally ignore the fact that women have agency. Sure women like money, but they choose on their own initiative to earn that money in different fields.

Your replies here are excellent in illustrating how DEI is more often than not a dog whistle for gender and racial discrimination. Women are already overrepresented at your company relative to the available workforce ("we're well above industry average in our woman:NB:man ratio"), and you've repeatedly insisted that discriminating on the basis of gender is legal. Your justification for this is that women make up less than 50% of tech workers.

But that's not how anti-discrimination law works. The law requires that applicants are discriminated against on the basis of protected class, not achieving equity with respect to the general population. If 80% of the workforce is one gender, then a non-discriminatory company would probably hire about 80% of said gender. Policies designed to push down representation of that group to bring it more in line with the general population isn't correcting for discrimination, it is actively perpetrating illegal discrimination. And that's what DEI is usually about.

1. https://en.wikipedia.org/wiki/Gambler%27s_fallacy

2. https://www.dol.gov/agencies/ofccp/faqs/AAFAQs

3. https://nces.ed.gov/pubs2020/2020167.pdf