Hacker News Reader

Comment by Kiro | Hacker News Reader

Am I the only one who think it's perfectly logical that the speaker ratio is the same as the ratio of the tech population? Of course we should try hard to find minority speakers just to be proactive since we do have a diversty problem. However, when you go to a conference you want to see the best speakers and then it's only logical that a minority is less represented since, you know, there are less good speakers to choose from.

Nordic.js is doing a great job with their "50% women speakers" policy but if we forcibly try to overrepresent every minority where do we end up? "50% black trans women speakers"? I wouldn't go to such a conference since the probability of bad talks would be too high.

View on HN

> but if we forcibly try to overrepresent every minority where do we end up? "50% black trans women speakers"? I wouldn't go to such a conference since the probability of bad talks would be too hig\

You end up with hiring quotas in companies, where a less-than-competent candidate can be hired instead of a competent one because he/she belongs to a minority. Will the bet on minority over competence will yield greater gains in the long run? That is a question for the ages. :)

That point you've apparently missed is that current hiring practises among some companies can be discriminatory. There are hiring managers out there who won't hire a female developer because they're just outright sexist. Quotas get introduced because there are brilliant minority developers who get passed over because they're in the minority.

All that said though, I don't believe there's any actual evidence that people end up hiring less-than-competent people because of quotas. It's only a theoretic possibility that could be true if the available candidates only included one incompetent person from the minority group, but I don't think that's ever going to be the case in practise. The choice always comes down to two or three competent people, and the quota system then pushes the decision on one direction. White male programmers don't like it because it's not in their favour, but the reality is that they could be passed over even if there wasn't a quota system. While people continue to discriminate quotas will be necessary.

the hiring process is completely opaque, so this can never be proven or disproven.

ideally, we'd have a system where we have a balance sheet that gets tallied up- but interviewing is an important part in gauging competence and culture fit (yes, this is a real thing, I know some developers who don't wash and I would not call them culturally fitting), so it's not something that can be solved easily.

We all, inherently have a bias- we can't divorce ourselves from that bias, so how do we hire? by tribunal?

I certainly believe quotas to be a shitty solution.

Inclusion like in speaking at tech conferences for one...

that doesn't change the fact that quotas are a shitty solution, perhaps having people vote publicly on topics that are on an agenda, while witholding the author's name?

quotas lead to inclusion. and inclusion is what's important. inclusion includes being represented at conferences.

content is what is important. technical content and content of character.

    "I have a dream that my four little children will one day 
    live in a nation where they will not be judged by the 
    color of their skin, but by the content of their 
    character."

How about running conferences on a completely blind process where the identity about the potential speakers are obscured by hashing, and people instead choose which speakers will present based on abstracts of the topic they will present on and some indication that they are qualified/knowledgeable to speak on the topic.

The best solution for inclusion is by eliminating the privilege through a blind process.

>Quotas get introduced because there are brilliant minority developers who get passed over because they're in the minority.

Any evidence (other than anecdotal) for this claim? The only reason I have ever seen for quotas is that the overall distribution is skewed, and therefore it must be discrimination.

Companies primarily hire out of networks/referrals. If your company is full of white guys, and their friends are very likely to be white guys, you end up hiring white guys.

Common sense would tell you it does happen. It only takes one guy who'd rather hire a man than a woman and one qualified woman applying for that job.

I don't think it's a meaningfully large part of the explanation for the skewed distribution. I think there are deeper systemic issues for that. However, I don't think I've seen more than a handful of people say that quotas are in place because of discrimination either. From my experiences, I think my view here is by far the dominant one -- there is overt discrimination, but the bigger factor is more subtle. It's primary and secondary education, culture, socioeconomics, marketing and advertising to kids, in short, a bunch of stuff that serve to skew the pool of people even getting to the point of wanting tech jobs. Quotas aren't aimed at getting Klansmen to have to grit their teeth and hire a black guy. Quotas are an attempt to bootstrap the system -- maybe if we just dump a larger than normal population of minorities into our little world, it'll self-propagate through more minorities seeing this as a viable option, with the hope that a generation or two from now, you'll have the distribution you want naturally.

That's a good list, but I think you have left out subtle discrimination. Things like the IAT show that there's a great deal of implicit racial bias in the US. [1]

I also have never heard of a company actually having quotas. What I have heard is people gathering statistics and saying, "Hey, this looks pretty bad. What are we doing that might cause this? What can we do to fix our [generally unconscious and unnoticed] biases in our hiring processes?"

[1] http://www.washingtonpost.com/news/wonkblog/wp/2014/12/08/ac...

You're right, I haven't heard of actual quotas either. I just got a bit loose with using the parent commenter's terminology.

I agree on the subtle and implicit bias issue as well. I wasn't trying to be exhaustive, but that's a big enough issue that it ought to be specifically mentioned.

There is plenty of evidence for hiring discrimination of all sorts. Oceans of it. For example:

http://www.nber.org/papers/w9873

http://www.nber.org/papers/w5903

It happens in all sorts of industries and across a variety of axes. I've heard some argue that no, tech is different, that we're special, special snowflakes. But they have no evidence to back that, just feelings that bias is something only other people do.

2:1 hiring advantage for women in STEM

http://www.pnas.org/content/112/17/5360.abstract

That study has many flaws especially in self selection:

http://othersociologist.com/2015/04/16/myth-about-women-in-s...

As newmana mentions, there are issues with that study:

http://othersociologist.com/2015/04/16/myth-about-women-in-s...

And there are other studies I find more persuasive. E.g.:

http://www.pnas.org/content/109/41/16474.abstract

http://blogs.scientificamerican.com/unofficial-prognosis/stu...

In particular, I think the "switch the names" sort of study under realistic conditions makes it clear that bias is an issue. And it certainly matches widespread anecdotal opinions of nondominant groups in STEM fields.

The authors address the criticism here: http://www.huffingtonpost.com/wendy-m-williams/women-scienti...

Most of this criticism doesn't stand up to even a cursory reading. For example: "not only are there less women in STEM fields [..]" No shit, Sherlock, I don't think anybody disputed that.

Or the study that shows a disadvantage for mothers vs. fathers. The very same study shows a hiring/renumeration disadvantage for childless men vs. childless women. So yes, there's a dynamic here, but it's not male/female.

Also, the original comment I replied to was that outstanding developers are overlooked. None of the studies I've seen that show gender/racial bias (there are a few), deal with outstanding candidates (the Wiliams/Ceci did, IIRC). And incidentally not sure why you refer to "studies", your two links refer to the same single study.

There was another study (I think it was a different one) touted as showing hiring biases in STEM, except it turns out the study was performed in psychology departments.

Etc. etc.

There's some great information out there on the web in case you are interested, for example the Pinker / Spelke debate. Here's a quote:

"PINKER: But that makes the wrong prediction: the harder the science, the greater the participation of women! We find exactly the opposite: it's the most subjective fields within academia — the social sciences, the humanities, the helping professions — that have the greatest representation of women. This follows exactly from the choices that women express in what gives them satisfaction in life. But it goes in the opposite direction to the prediction you made about the role of objective criteria in bringing about gender equity. Surely it's physics, and not, say, sociology, that has the more objective criteria for success."

http://edge.org/3rd_culture/debate05/debate05_index.html

There are hiring managers out there who won't hire a female developer because they're just outright sexist.

Do these people actually really exist in 2015, and are there organisations that would allow them to behave like that? I've known quite a few hiring managers, and some of them have definitely made stupid comments about "how nice it'd be to have a hot girl on the team" (though equally they might also make stupid comments about hiring graduates - of either gender - to make cups of tea), but none of them, to my knowledge, would ever have not picked a female candidate who was technically up to the task purely because she was a woman.

Have I just led a sheltered existence?

(I guess I'm talking here about SMEs and larger companies; I suppose if you're a tiny company run by a couple of sexist pigs, then any kind of behaviour is allowed to stand.)

It wouldnt be tolerated if it was spoken outright. Unfortunately its pretty easy for a hiring manager to dismiss someone officially because of something on her resume or in the interview officially, while being sexist or racist, or just having embedded biases in their head. Ive heard it often enough.

Anecdotally, yes I've met people who won't hire a woman. Reasons vary from an assumption that they won't be as capable because they have interests outside of coding, a belief that other developers won't be as productive if there's a woman on the team because they'll be distracted, or that they won't want to be a part of the company long term because they'll want to leave to have a family.

> That point you've apparently missed is that current hiring practises among some companies can be discriminatory.

Yes, I know they are. My question ("Will the bet on minority over competence will yield greater gains in the long run?") was genuine; I really wish I knew if positive discrimination to counteract the existing discrimination will improve the situation.

> All that said though, I don't believe there's any actual evidence that people end up hiring less-than-competent people because of quotas.

Evidence? Maybe not. But believe me, there are quotas with perfidious results in large companies. I've seen it happen somewhat like this: department X has 5 vacancies to fill; due to department "demographics", company mandates that 3 of the 5 vacancies are for female candidates; due to candidates "demographics", department X ends with a pool of N candidates for 3 vacancies and another pool of 10N for the other 2.

But, as you said, it's theoretical. If the candidates had a quantifiable "skill", the "no subpools" approach would yield better results - a bigger "skill" quantity for the total of vacancies. However, in practice, you can't quantify skill, and much less the impact a candidate will have in the department. The apparently sub-optimal "subpools" approach might get the 5 vacancies with "more than enough" skill; The apparently sub-optimal "subpools" might even yield better results due to e.g. diversity. That's why I said that this is a question for the ages - only cohort-like analysis of hiring practices will give us something close to a answer which isn't merely speculative.

The choice can't possibly be "always between two or three competent". Less competent exist, and they have to pop up somewhere. With quotas, once most competent from the lesser set get taken, only less competent will remain available. It's a trivial mental exercise, not something that requires gathering evidence to figure it out.

    That point you've apparently missed is that current hiring
    practises among some companies can be discriminatory.

Totally agree that current hiring practices have implicit biases that can be discriminatory. What I don't get is how people can consider alternative hiring practices that are explicitly discriminatory okay, because they are discriminatory in the opposite direction.

It would be nice if people applied some heuristic like the "5 whys" and tried to actually perform a root cause analysis and try to resolve the base causes instead of plastering over the problem with practices that are at best no worse than current practices.

Explicit reverse discrimination is still discrimination. Quotas result in a situation where brilliant majority developers get passed over because they are in the majority. How about we try and come up with a solution where equally brilliant developers regardless of their minority or majority status don't get overlooked?

The fear of not having the best people speaking at a conference might have some merit if the people speaking at most conferences today were actually the best people. They are usually not. The best people speak at very few conferences because they are busy working.

In other words, women are overrepresented in conferences compared to their presence on the field, and it's going to be the same with ethnicities. In my business experience, women are already overrepresented in management positions compared to developer positions. Who wants to remain a developer anyway?

> Who wants to remain a developer anyway?

I do.

Just my opinion, but the idea of "50% women speakers" isn't out of some sort of over-representative fairness to a "minority", but a reflection of real world demographics.

In any case I can think of, roughly 50% of people in a given conference's catchment area are women, and since tech is an industry in which anyone could excel, having "50% women speakers" is promoting and encouraging a desired ideal to both the industry and potential participants in that industry. (I must note that "failing" at having 50% women speakers at a conference is by no means a fatal flaw, but having the ideal there and the principles in mind can still effect long term change.)

The idea of having 50% "trans" people at an event would be a very different one given less than 1% of people are trans, allegedly (although given the stigma still attached, I think being trans needs to become more acceptable in society as a whole before we'll know the true number. They're probably a lot higher than we'd suspect, as we saw with homosexuality through the 20th century).

> "50% black trans women speakers"?

Bingo, this seems the natural end (or just the beginning of absurdity, really) of emphasizing equal representation of intersectionality, and it doesn't seem productive, dignified, or fair. Can anyone show that this is distorting the article's point? I don't see why transexuality should be minimized more than race or gender, and why it's not a component of intersectionality. Or defend why it's actually a good idea to turn people into a tuple of (gender, race, ethnicity, age, sexuality, sex, sex at time of birth, religion, caste, language) and select uniformly (or oppression-intersectionally)? How are you even going to find all those things out? (In some cases, asking would be a legal liability, and doesn't seem the most appropriate thing to ask.)

I'd rather wear down barriers where I see them and let diversity organically grow, than to constantly be categorizing people and fudging the numbers into something artificially diverse.

> if we forcibly try to overrepresent every minority where do we end up?

I am so tired of this sort of fantasy catastrophe based on a fantasy exaggeration of of a reasonable point.

If your question is, "What would happen if we reversed the polarity on the diversity-suppression machine and suddenly conferences were as diverse as they had been previously vanilla?" then the answer is: the conference would be about the same, but with a different set of speakers.

Heck, let's try it. The industry has been unusually straight white male for, what, 30 years? [1] Let's try 30 years of conferences that are exactly as biased but in the opposite direction. In fact, let's just try 3 years and see how it goes. My guess: lots of good conferences and a much more diverse set of people joining the field and not quitting the field because they're tired of dealing with bias.

> I wouldn't go to such a conference since the probability of bad talks would be too high.

You understand, don't you, that your implicit assumption here is that there just aren't enough competent non-straight-white-dude speakers to fill out a conference list?

In my experience, the conferences that work really hard on diversity are better. Why? Because to get a diverse speaker base, you have to work really hard on outreach. You're going out beyond the usual suspects. You're finding people who wanted to speak but had never got a shot. You're finding the people who never even thought about speaking but would be great.

[1] http://www.npr.org/sections/money/2014/10/21/357629765/when-...

    You understand, don't you, that your implicit assumption 
    here is that there just aren't enough competent non-
    straight-white-dude speakers to fill out a conference 
    list?

I think you're deliberately misrepresenting the OP. The way I read the OP, they are making the assumption that for every single identity group in tech, that the prevalence of great speakers to bad speakers is approximately equal. For shits and giggles, let's say that assumption is 2% of white guys, 2% of white women, 2% of black women, 2% of black men, 2% of asian women, 2% of asian men, 2% of trans people, etc. etc. etc. are no only competent by talented speakers you'd want to see at a conference.

The moment you intentionally ignore base rates of the availability of each group, you're no longer getting the 2% of the engineering population that are the best speakers regardless of their identity. You're instead optimizing for diversity over technical excellence. Nothing wrong with that, just be clear that you can't optimize for both unless the base rates for all groups are approximately equal or unless one identity group actually have a greater percentage of technically excellent speakers than another identity group.

That said, I don't really care about conferences either way, just statistical truths. I rarely go to conferences any more, since I prefer reading code, papers and technical writings that are shared based on the quality of content instead of the identity of the author.

    In my experience, the conferences that work really hard on 
    diversity are better.

I'm really curious to know your definition of "better" because I would venture that both you and the OP are completely correct in your viewpoints and arguments, because you two don't share the same definition of "better". Those conferences are better based on your criteria of better, but may be subjectively worse for the OP based on their criteria of better.

FWIW, I'm a non-straight "white" dude (half anglo, half latino (by way of iberian and mediterranean descent) (more specifically a anarchoamorous, pansexual, agender male with a cisgender visual appearance because I don't see the value of wearing my identity on my sleeve)

> you're no longer getting the 2% of the engineering population that are the best speakers regardless of their identity.

That's not what you're getting now, either.

And note that you've shifted the goalposts. He was not claiming that the talks would no longer be the world's best possible speakers. His concern was that the "probability of bad talks would be too high".

> Those conferences are better based on your criteria of better, but may be subjectively worse for the OP based on their criteria of better.

All sorts of things could be true. Unless you're claiming that they actually are and that this difference is material, it seems irrelevant.

    That's not what you're getting now, either.

Completely agree. The point I was making is that a quota system not only isn't better, but is worse if those quotas are not proportional to base rates of the population you are drawing from (software engineers in this case).

The only way to select speakers based on quality of content is to select speakers by content alone and hide all other criteria and characteristics of the speakers when selecting. Pretty much any other system introduces implicit and explicit biases, especially a quota system.

FWIW, yes, I'm claiming that your definition of better is different than that of the OP and that this difference is material. I may be wrong about this but based on your arguments in this thread and their arguments, if feels to me like you're both optimizing for different things. You're optimizing for diversity and the OP is optimizing for quality of content. The only way you get both is if the population from which you're drawing speakers at random has already been optimized for diversity.

There are 100 fibbwobblers. 10 are blue, 90 are green. Approximately 10% of fibbwobblers are great conference speakers (evenly distributed among all fibbwobblers regardless of color).

You want to hold a fibbwobbler conference with 10 speakers. To get only good speakers, you'd need to draw the 1 out of 10 blue fibbwobblers that are great speakers and the 9 that are green. If you have no quota system and end up with 10 green fibbwobblers speaking, approximately 1 of them is likely going to be bad speakers. If you decide to have a quota system where fibbwobblers of both colors are equally represented, 50/50, then you'll likely end up with the one great fibbwobbler speaking that is blue, and 4 blue ones that are not great speakers and 5 green ones that are great speakers. Under no quota, you get 1 bad speaker out of 10, under and equal representation quota system you get 4 bad speakers out of 10. Keep in mind that this is all based on the perfectly reasonable assumption that all fibbwobblers are equally capable of being a great speakers, regardless of their blueness or greenness.

Your worst case scenario in terms of quality of speakers gets comparatively worse as the spread between % of population that has identity X and the % of speakers that must be of identity X widens.

This is root of exception the OP was pointing out.

> quota system [...] You're optimizing for diversity and the OP is optimizing for quality of content.

I have never advocated for a quota system and I am not optimizing for diversity. Perhaps you should go argue with someone who holds those views, as I find this irritating and tiresome.

> There are 100 fibbwobblers. 10 are blue [...] 10 speakers [...]

And here you construct a ridiculous scenario to generate false intuitions. In particular, you are trying to generate the exact false intuition I pointed out with my first post: the notion that "there just aren't enough competent non-straight-white-dude speakers to fill out a conference list". It's horseshit. Peddle it elsewhere.

> I wouldn't go to such a conference since the probability of bad talks would be too high.

Why?

Because only a small % of all people in tech are good speakers. Having a smaller population to choose from would mean a higher risk of putting a bad speaker on the lineup.

You're assuming that a) good-speakerness is a fixed quality, and b) that the speaker selection process is something like pulling bingo numbers. Neither is true.

Reply on news.ycombinator.com