Good way of putting it. I think most people would prefer equality of both treatment and outcome, if possible.

Additionally, I suspect that it is possible, just not in the short term (or even decades, possibly). But in the long term, I think that if you achieve equality of outcome, then new norms begin to take root, and eventually, you can remove any artificially created inequality of treatment and still get an equal outcome.

Your function example really nails it down if you consider it cyclical, with the output being averaged with previous inputs and then re-entered as input. So, inequal input to an equal function generates inequal output. But if you change the function to be inequal but opposite to the existing input, then your output becomes equal. When you average it, it becomes inequal again, but less so than the original. Once your average approached equal by repeating this, you can replace your inequal function with an equal one and your system should remain fairly stable. Essentially, your function just needs to compensate for biased input, and if there is no biased input, it can behave equally.

However, simply having equal treatment/function with inequal outcomes/input/output results in a system that just remains the same.

So basically, affirmative action type steps could potentially lead to an environment where those steps are no longer necessary, and can be removed, achieving actual equality.

But maybe I'm an idealist.

I don't know if you're an idealist, but you seem to believe that there is strong feedback in the function, which is something that I am somewhat skeptical of.

If income and examples/role models were the primary inputs, I'd have to agree that your feedback model would work. The problem is that there seem to be fundamental differences between many groups in IQ, EQ, and preferences. In addition, men and women have different life experiences beyond how society treats them, as men cannot experience what it is to carry a child to term, and men are less likely to breastfeed their children; if this type of experience is a relevant input to work productivity or preferences, a 'fair' system will probably create unequal outcomes.

Perhaps there are some factors that would lead to unequal outcomes that would not be corrected, but I expect that these, including the ones you describe, would have significantly smaller effects than what we currently have. And I expect that those would have been the initial impetus causing a runaway feedback effect which would cause what we have.

So, really we have two functions. One function produces the number of men and women applying for the jobs. The second function produces the number of men and women occupying the jobs. The second function is the one we've discussed wanting to control to be either fair or unfair. And yes, with inequal inputs, a fair function produces inequal outputs.

The first function, we are assuming for now is beyond our control, and the inputs to it are largely unknown and/or beyond our direct control. It is a product of experiences, societal norms, encouragement, discouragement, and personal interests. The first function directly generates the inputs to the second function. As I understand it, you are postulating that this first function is inherently inequal to the point of generating the current applicant proportions we are seeing, and that the output of this function is not significantly impacted by any inputs that are the output of the second function (feedback effect). I would argue that the first function, while possibly inequal, is not sufficiently so to create what we observe, and instead, is reasonably impacted by the output of the second function. Thus, the feedback effect.

Anecdotes I've read and witnessed seem to indicate that there are factors that reduce the number of females who make it to the application stage that would not exist if females were not already a minority in the workforce. Males in an all-male team do frequently develop habits that would put off most women who join that team as a newcomer. This doesn't have to be intentionally malicious to discourage a woman from continuing to work in the field. If women are made to feel uncomfortable, intentionally or otherwise, they are less likely to stay and to encourage other women to join, and more likely to actively discourage more women to join. And it seems very likely that a 50% female workforce would go a long way to removing uncomfortable work environments. And this can happen similarly at the college level before ever getting to real work environments, with women facing friction in mostly male programs.

Then there's the conception that it is a typically male job, just because there are many more man than women currently in it. This isn't malicious either, necessarily, it's just the way human minds work. You see mostly men doing something, you associate it with men. So you get a stronger unconscious gender affinity just because that's the way it is and has been. And that affinity leads fewer women to try the field out, fewer people to encourage them to try it out, and more people to encourage them toward other fields or even some people actively discouraging them, possibly those who have previously dropped out of the field or possibly just naive types.

I'm sure there are other examples of ways that an existing imbalance in the workforce impact the balance of those attempting to join the workforce and making it all the way to the application phase.

Also note. I'm not making any claims as to exactly how strong this feedback loop is. Just that it is strong enough to make a significant difference over a significant amount of time. That time frame may be decades, or it may be generations. I expect that at a minimum, we'd see a far greater natural equality if it were artificially enforced for 50 years or so. After that, you could have a system that is almost entirely fair, perhaps occasionally adding a small balance to restore it any time it approaches a significant tipping point again.

My personal view is that the case you are putting forward is quite tenuous, and does not justify unequal treatment. Making a "significant difference over a significant period of time" is simply too speculative to justify discrimination, especially when the "significant difference" wouldn't necessarily increase people's happiness, productivity, or life satisfaction; one could imagine that making men and women more similar might make both groups worse off. In addition, it seems that most groups' positions and roles seem to be evolving (I assume mostly in their favor), and I have no reason to believe that spontaneous evolution is any 'worse' than affirmative action, especially in long term results.

To make your case convincing to me, you would need a clear outline of how the plan would create a Pareto improvement (which would not necessarily be pecuniary), and you would also need to outline a way of determining whether the plan is working. In my view, there is a danger of executing a plan which has no chance of working, and continuing to rachet-up the discrimination in the vain hope that 'it might work if we try harder'. I would also require your plan to have a much better expected outcome than the status quo.

Perhaps there is also an alternative argument which would convince me that procedural discrimination was 'worth it', but I haven't heard it yet.

Well, money doesn't buy happiness, but it does go a pretty long way toward facilitating it. So I'd say it's unlikely that the status quo of leaving a huge imbalance in some of the most lucrative fields available today wouldn't be worse for women than addressing that imbalance. I can't think of any high paying fields that are dominated by women. And there is a notable difference in average pay between men and women.

I'd agree with your second paragraph, for the most part. It is easy for a program with the right intentions to fall into that trap. It would bear a more in-depth analysis. I think, once started, it would be fairly easy to determine if it is working. If a more equal workforce can't be maintained without an increase in efforts, it probably isn't working. In fact, a sign of it working would be a gradual decrease in the need for it. In other words, if the workforce began to skew the other way, then the program would need to be scaled back to bring things back in line. So, if it worked, it would be self-terminating, and should be designed as such.

>Some people prefer to have equality of treatment, while others prefer equality of outcome; I think this is why people disagree so vehemently.

Yes, because those who prefer equality of outcome are wrong, and they see no problem with discrimination against me when it comes to education and employment. In order to bring about equality of outcome along gender or racial lines you would also have to create separate job markets along those same lines. Instead of competing against everyone for a position I would only be competing against white males. At some point in our history it was decided that separate is by definition not equal.