Exactly. And that's what the pipeline argument tells. Whereas the more vocal pro-diversity movements tend to assume that if you haven't fixed that problem directly and immediately, it must be your moral failing.

> Whereas the more vocal pro-diversity movements tend to assume that if you haven't fixed that problem directly and immediately, it must be your moral failing.

I don't know about that interpretation. I think a better interpretation would be that diversity advocates would say "if you haven't fixed the problem yet, you should continue to work to fix the problem, even if it takes more time and effort than you initially hoped."

I don't understand what you mean.

Are you saying that the people who are receiving a biased mix should introduce an opposite bias to counteract it before sending on to the next step, or not?

Possibly I am misunderstanding what a "pipeline problem" refers to.

Wow, I'm experiencing some deja vu. I feel like I already (like, a year or more ago) sent this comment and got a reply, but I don't remember what the reply was.

... Ok, I think this is the thing that doesn't make sense to me: >> wherever you try to improve the pipeline, you cannot expect to make things better than they are at the input of your pipe segment.

> But that's basically just complaining that there's no quick and easy fix to your hiring problem.

I don't understand this response. Saying "It is impossible to do x without doing y or z", saying "That's just complaining that it is hard to do x without doing y or z. Tough, that's what you have to do." doesn't make sense.

Responses that could make sense include "It actually is possible to do x without doing y or z, and that is what you have to do." or "You have to do x, even if it means doing y or z, even though doing y or z is hard".

Were one of these forms what you were saying? (Note : it is very possible that I simply misunderstood what you meant. This is the thing that I think I will probably conclude if you respond.)

The options that I can see for hiring people are: Introduce an opposite bias when hiring in order to counter-act a bias in the input Don't do that, simply work to not introduce any additional bias, and accept a biased result in the output Somehow(?) act in a way that works to make it so that the input received is not biased (or options worse than any of these)

If you are counting "the input" as meaning "the set of people who apply" then I suppose you might be able to influence "the input" by changing where you advertise the position, or things like that, but that seems to me like "potential applicant -> applicant" is just another step in the pipeline, and if the inputs there are also biased, then the same trilemma applies.

If one removes all the bias potentially introduced from each step which one can directly control, then one can no longer directly remove introducers of bias.

And I don't think smalltime company TUV's role is to work to change educational outcomes and the like, and therefore if this and its predecessors are the only remaining sources of bias, I don't see how it is their problem.

That's not to say that those are usually the only remaining sources of bias. I wouldn't be surprised if many such companies do introduce bias which they can and should address, but if the claim is that they should produce an unbiased output despite having a biased input which they cannot influence, then the only thing that can follow is that they should introduce a bias in the opposite direction.

My point is that, if there is in fact a pipeline problem, that probably does not constitute a shift in responsibility away from the hiring company. Saying "you cannot expect to make things better than they are at the input of your pipe segment" is explicitly attempting to shift responsibility away from the hiring company. I simply disagree. I think a company is still largely responsible for the outcomes of its hiring process, even if fixing those outcomes turns out to be an extremely hard problem that cannot be solved overnight with a diversity webinar.

Are you saying:

1) the input is not actually biased

2) the company is responsible for making the output less biased than the input (or produce an unbiased output from a biased input)

3) the company is responsible for making sure that they don’t cause the output to be more biased than the input, (but you are not saying that it is responsible for more than that)

3’) The same as 3, except that the amount of bias in the input, while nonzero, is small and relatively unimportant

4) The company is responsible for preventing the bias in the input

5) I (drdeca) missed an option; none of the above (if so, please specify the option I missed, which can be the case while none of the above are)

My leanings are that either (2) or (3) (or 3’, but that is just a subtype of 3) (or 2’, but again that is just a subtype if 2) is the case.

With my current understanding, I believe I can answer this one. It's #2. Specifically, your parenthetical comment in #2.

Bias exists and is a problem. It is worth spending additional resources to draw additional samples from the underrepresented population in order to offset that bias. With more samples from that population than the previous pipeline stage would provide if you sampled evenly, you can counter some of the bias without adjusting any standards of quality (capability) based on the population being sampled.

If you are a member of the overrepresented population, you will have a smaller chance of being hired with this intervention. But for someone who is hired, their skill level will be independent of what population they belong to. (Women will not be given an easier interview. They will be more likely to be offered an interview in the first place.)

Whether that is "fair" depends entirely on how you define fairness. This procedure stacks the odds against a man. Our current overall system stacks the odds against a woman. Both can legitimately complain about unfairness.

Thus, it's largely irrelevant that the percentages in earlier stages of the pipeline mean that there's no way for the overall balance can reach 50/50 through only later-stage interventions. Where did this magic 50% figure come from? The only point is to improve the percentage from where it is now. The issues that prevent achieving 50% are real, but don't prevent progress anywhere in the pipeline.

I imagine the additional costs of sampling more from a smaller population would rise dramatically the closer you try to push the outcome towards 50%. So companies will have to decide how much they're willing to invest. Fortunately, there is still real value in pushing beyond the status quo even if you don't get to 50%.

(More generally, the magic number is not always 50%. It's the proportion of the URM in the overall population. 50% is roughly the female part of the population in areas advanced enough to have these sorts of jobs available.)

thank you, that seems like a plausible idea to me.

It also seems possible to me that such an intervention might reduce feedback loops which cause the bias in the input anyway. Like, if women who are aware that a smaller proportion of the people employed in a field are women than the proportion of people in the population, maybe they might see that as a factor weighing against picking that field as one to go into? (of course, some might find it a reason to pick that field. I just mean that it seems possible that there is such an effect on average), and so, the #2 intervention might reduce such a feedback loop problem, if such feedback loops do exist.

Right. And those feedback loops do very much exist -- you can find plenty of women sharing their experiences of what it feels like to be the only woman in a room full of men. Over and over again. Being an "only" is rough. I'd imagine that going from 2 to 3 in a group is a much, much smaller change.

It also means that success often requires acting like a man, even when that is neither natural nor optimal for the situation. That's one reason why more than minimal diversity helps -- if there is a benefit to be had from diversity, you may not get it from having an "only" who is pressured to fit in.

I'm saying that, if one thinks diversity is important [0] at companies, then the outcomes of hiring and retention are largely (overwhelmingly) the responsibility of the hiring company. This is especially for very large hiring companies.

[0] If you don't think diversity is important, then that's another discussion altogether. In this thread I have been operating under the assumption that we agree it is important (and in fact I personally do).

That means that in my view it's just not reasonable or acceptable for a company to say "diversity is really important to us, but we tried some things like diversity training for our hiring managers, and that hasn't worked, so we looked up some statistics and we have concluded that it's a pipeline problem so we can't be expected to do anything about that."

I was also working under the assumption that diversity was important, or, at least, that a bias is bad.

I don’t see any conflict between this and the list of options that I listed.

Do you think that my list of options is missing an option, or are you refusing to pick one?

If the latter, why are you refusing to pick one?

Do you think that the list isn’t logically valid?

I am asking simple questions which I am aiming to make as easy to answer as I can, and you seem to be dodging the questions, and are instead responding with an accusation.

It is a simple question, and I left a clear option for in case the question would otherwise accidentally presuppose something false.