Hypothetical affirmative action scenario

[Loren Pechtel]

I can't think of a better recruiting tool for the KKK.

Really? I heard they have good retirement plan benefits.

How about addressing the point?

 

[ronburgundy]

A productive exchange as usual.

The question of whether could be used for net positive effect is more complicated.
Since all racial discrimination, regardless of motive, inherently leads to less qualified hires, the quality of work done by the government and police would decline, causing harm to both blues and greens impacted by these officials. Plus, it would legitimize and therefore increase racial discrimination generally by eroding the principle that such discrimination is wrong, which would particularly harm the blues since they are a statistically minority. OTOH, even though the additional blues hired under such policy would be less qualified, they might have more motive to counter the presumed racism being engaged in by the greens in those positions. Plus, they would serve as examples to other blues to strive to become capable to enter those positions w/o the need to be helped by the affirmative action policy.

Thoughtful overall analysis (your post I mean) with which I would not take much issue. Regarding the above specifically (for which I have omitted the last part about college applications because it's not really the OP scenario) I might also add that in a 'divided society' where one group are or feel they are second-class citizens, the quality of the service can improve because the service (eg police) is seen to be more of an equitable representation of the population it serves, which effectively can make the job easier to do, and the functions of such services are to some extent a two-way street between the provider and receivers. This is distinct from whether the performance of individuals (or indeed the police force by extension) declines because of a lower entry bar.

True, and I thought about mentioning that, but my post had already introduced too many variables. OTOH, if that is going to be the justification, then the "blues" should be assigned specifically to serve the blue community. I wonder how comfortable AA advocates would be with that. Also, if the blues in the public have a legit objective reason not to trust the non-blue service providers that is one thing. But if it's just a form of racist distrust itself, should we be reinforcing the notion that you can only trust someone of your same race? It certainly would limit the context and type of jobs the policy should be applied to (maybe cops but not teachers). You'd need to make the argument that being the same race as the person being served is itself a "qualification" to do the job well.

I am also at least a bit sceptical of whether lowering the bar does in fact necessarily mean lower performance of the sort you mention in any case.

If the qualifications being used have any empirical correlation with performance, then less weight on those qualifications mathematically means lower predicted performance (not in every instance, but in the aggregate). Again, you would have to assume that race itself is job impacting qualification for the inclusion of race as a factor not to have a net negative impact on performance. Obviously, if the qualifications have zero correlation with performance, then given them less weight would not predict lower performance, but then they aren't really qualifications.

As I often cite, we had very strong AA quotas here for getting in our case 'greens' (aka Catholics) into the police, which I think was 98% Protestant at its height. I think I am right in saying that the AA had a target of 50% for each, though it did not get to that before it ran its 10-year course (I think it got to 30-something% for catholics) And as far as I am aware, there was no loss of quality in performance that was reported as being measured.

Taking those as facts, what they logically imply is that prior to the policy implementation Catholics were being selected against despite Catholicism having no correlation with qualifications that predict performance. Thus, the policy merely reversed the direction of this existing factor, and since is was uncorrelated with performance that reversal had no impact on performance. So, yeah, if race or another demographic factor is already being used to discriminate w/o any correlation to performance-relevant qualifications, then reversing that discrimination will not lower performance. However, the far better solution is to removing the discrimination against Catholics rather than reverse it. This would mean that only performance-predicting criteria are used, which if Catholicism truly has no job-relevance would increasing Catholics w/o having to reverse discriminate in their favor, and would increase the level of performance compared to either the anti or pro Catholic discrimination policies, b/c now there would not be non-performance factor reducing the weight given to performance factors.

It would be surprising if over all examples there were no failures. How much they could be attributed to the AA is another matter.

I would argue that simple logic and math necessitate that AA would be the cause of lowered performance in every situation, except those situations where the the demographic factor being used was already being used in the reverse AND has no empirical correlation with performance-relevant qualifications. I would argue exception those situations are rather rare, b/c if the group is already being discriminated against, then odds are it has been discriminated against in other ways that would lead members of that group to be less qualified, as is the objectively the case with race and most positions that are impacted by academic/intellectual development. In those far more likely scenarios, simply using qualifications would lead to under-representation of the group, and thus adding group-category as a deciding factor favoring that group would inherently reduce the average level of qualifications of those given that position.
And again, even in those exceptional situations where reversing an existing direct discrimination would not lower quality, it would still produce less equality (and more injustice, resentment, and promotion of bigotry), than implementing neutral policies were clear cut objective job-relevant qualifications are used.

Another factor could be automatic resistance to the implementation of the AA. Sometimes this can be called a backlash. A backlash can, of course be justified, and/or happen 'for the best of reasons' but I suspect in this sphere of activity it often doesn't.

Given that the AA were are talking about C thru F, is by definition discrimination against particular groups that devalues individual merits, I would say that the "backlash" is ethically required by anyone who thinks discrimination that ignores that devalues the individual is wrong and at odds with core progressive values. Plus, most of the people in the majority group negatively impacted by such policies will be those who are below average in advantages overall, b/c they are the one's closer to the cutoffs who are the first to cut rejected when being in that group is weighted negatively. Bear in mind that C thru F logically entail using different cutoff criteria for people of different groups. Mathematically, that is what using group membership as a weight in the algorithm means. It lowers the criteria needed for the minority group and raises the criteria for the majority group. So, it is those near the cutoff who get cut when their criteria for their group is raised. So, even people who don't have a principled moral concern about individual-devaluing discrimination, have a reasonable and rational basis to object to policies most likely to cause them or their loved one's harm. Plus, those most harmed are those with the most in common with the minority group in terms of SES, and thus this resentment undermines what should be a socio-political alliance.

 

[ruby sparks]

True, and I thought about mentioning that, but my post had already introduced too many variables. OTOH, if that is going to be the justification, then the "blues" should be assigned specifically to serve the blue community.

You might not have to take it that far. It might even be counterproductive, as I'm sure you'd already have thought of. It might be better, and enough, to 'mix it up'. I would not, I suppose, rule out a slightly greater proportion of blues policing blue areas. But one would have to be careful, I think.

To wander briefly away from the OP scenario(s), there is a similar issue with other jobs, such as for example doctors. It may be, for example, that blue doctors are more ready and willing to work in poorer 'blue areas'. As before, I think we'd have to be careful about endorsing this, but on the whole we might not prevent it. It's slightly different if the person doing the job can choose where they want to work, I suppose.

Also, if the blues in the public have a legit objective reason not to trust the non-blue service providers that is one thing. But if it's just a form of racist distrust itself, should we be reinforcing the notion that you can only trust someone of your same race? It certainly would limit the context and type of jobs the policy should be applied to (maybe cops but not teachers). You'd need to make the argument that being the same race as the person being served is itself a "qualification" to do the job well.

Yes, there would have to be some sort of legit reason why the blues wouldn't trust a green police force. Ditto for any job, I suppose.

My scenario does involve the disadvantages of the blues, as a group, being at least partly to do with unfair treatment by the greens, historically and to a lesser extent currently. I can't remember now if I said that explicitly in my OP. I certainly meant to imply it when saying for example that the greens had more societal privileges.

If the qualifications being used have any empirical correlation with performance, then less weight on those qualifications mathematically means lower predicted performance (not in every instance, but in the aggregate). Again, you would have to assume that race itself is job impacting qualification for the inclusion of race as a factor not to have a net negative impact on performance. Obviously, if the qualifications have zero correlation with performance, then given them less weight would not predict lower performance, but then they aren't really qualifications.

Sure. I suppose I'm saying 'what are the qualifications that are likely to be performance-enhancing for say the police or a governmental role?' Academic qualifications may not be one of them, for example, at least not beyond a certain threshold.

Can 'being blue' of itself be a performance-enhancing qualification. Probably not, but I have been talking about performance in the round, of the police force. So if I agree (as I think I should) that 'being blue' of itself confers no additional qualification at all, that's ok, because we can look at the overall picture and not the individual. I think that is actually the assumption made in such policies. I don't think anyone thinks that catholics, say, or blacks (on planet earth) are better because just they are, in itself, either of those.

Taking those as facts, what they logically imply is that prior to the policy implementation Catholics were being selected against despite Catholicism having no correlation with qualifications that predict performance.

It was partly that, yes. As ever, it was a bit more complicated. That, what you say, was happening, definitely. In tandem, and for more than one reason, catholics were not applying in large numbers, partly because of discrimination they expected to receive if they joined, partly because of intimidation they might face from catholics in their community and partly because at least some of them objected in principle to joining the 'oppressors'.

Thus, the policy merely reversed the direction of this existing factor, and since is was uncorrelated with performance that reversal had no impact on performance. So, yeah, if race or another demographic factor is already being used to discriminate w/o any correlation to performance-relevant qualifications, then reversing that discrimination will not lower performance.

Again, this 'performance of the individual' is one way to measure. I'm not sure if it's not slightly abstracting the situation.

Also, tangentally, what do we even mean by 'performance of the individual'. Would, 'a tendency not to be prejudiced' be one? If so, let's say hypothetically that a blue police recruit is less likely to be discriminatory, perhaps because he has been on the receiving end of it in his life and knows how unhelpful it is. I'm not saying this would necessarily be the case (it might indeed not be the case) but I use it as illustration. It's a bit hypothetical.

Actually, I myself do not know why having a race quota would not adversely affect the performance of the individuals who get in with less qualifications (whatever they may be). I would in fact intuitively think that there should be some loss of performance quality, in many jobs. If there were not, I might wonder why not. I offered one possible explanation, that the minimum requirements are raised across the board at the same time as implementing the quota. Another might be (for the police) enhanced training (for all recruits) after being accepted as a recruit. But I actually don't know if either of those was implemented here, for example. I do not rule out that possible decreases in performance took place, but were not reported, it not being in everyone's interest to admit them, especially not on the part of the authorities who implemented the quota, who would have a vested interest. That said, I myself have not noticed any decline in performance (whether because of the quotas or for any other reason) and I am not aware of any reports or suggestions/allegations of it, and there definitely plenty of those who opposed the quotas (protestant/Unionist politicians in particular, and many of their supporters) who would like to pounce on such a thing.

However, the far better solution is to removing the discrimination against Catholics rather than reverse it. This would mean that only performance-predicting criteria are used, which if Catholicism truly has no job-relevance would increasing Catholics w/o having to reverse discriminate in their favor, and would increase the level of performance compared to either the anti or pro Catholic discrimination policies, b/c now there would not be non-performance factor reducing the weight given to performance factors.

Hypothetically, perhaps. But in my opinion, if the imbalance is severe enough, and is the result of historical or ongoing unfair discrimination, then something more can be argued to be both necessary and better than simply levelling the application playing field. For example, had there been no AA here (and quotas were only a part of it) it is likely that change would have been too slow.

I think it very much indeed depends on the severity of the issues. That is why I think quotas are a measure of last resort only.

I would argue that simple logic and math necessitate that AA would be the cause of lowered performance in every situation, except those situations where the the demographic factor being used was already being used in the reverse AND has no empirical correlation with performance-relevant qualifications.

Logic and maths would strongly suggest it, yes, if we were only looking at the performance of the individual recruit. Perhaps human affairs are too complicated, and humans too capricious, for us to only use logic and maths in an analysis. What we would need, ideally, is some good evidence based on wide-ranging analysis. Here is a question; "does introducing quotas necessarily result in lower performance?" I don't know the answer. It might be a different answer for police forces than for, say, lawyers or doctors. Numerous follow-on questions can be imagined.

I would argue exception those situations are rather rare, b/c if the group is already being discriminated against, then odds are it has been discriminated against in other ways that would lead members of that group to be less qualified, as is the objectively the case with race and most positions that are impacted by academic/intellectual development. In those far more likely scenarios, simply using qualifications would lead to under-representation of the group, and thus adding group-category as a deciding factor favoring that group would inherently reduce the average level of qualifications of those given that position.
And again, even in those exceptional situations where reversing an existing direct discrimination would not lower quality, it would still produce less equality (and more injustice, resentment, and promotion of bigotry), than implementing neutral policies were clear cut objective job-relevant qualifications are used.

I'm not sure if I grasp your point there, but yes, there is the issue of whether specifically the academic qualifications attained by the applicant are or are not a good measure of their abilities. If, for example, a blue candidate had been 'held back' by, say, a poor education or other disadvantages, their qualifications might be artificially lower than for, say, a green candidate who not only benefitted from good schooling, but extra tuition. A hypothetical blue candidate with a lower score may therefore be 'brighter' than a green candidate with a good score. The blue applicant with the lower score may thus catch up and overtake, or simply be better at their job than, the green one, in performance terms. Similarly, if the blue applicant can demonstrate some quality such as 'having overcome obstacles', they might have more of this (potentially very useful) quality than a green student who has not had to overcome those obstacles. Or, perhaps a blue candidate will be more motivated to perform better, because of expectations upon them (for example that they feel they are or are seen as 'representing their colour' more than was the case for greens). There might, in other words, be a number of variables and factors beyond what are typically referred to as 'qualifications'.

Given that the AA were are talking about C thru F, is by definition discrimination against particular groups that devalues individual merits, I would say that the "backlash" is ethically required by anyone who thinks discrimination that ignores that devalues the individual is wrong and at odds with core progressive values.

If someone thinks that, yes, but I myself would not go along with it, because I would not subscribe to the notion that all forms of racial discrimination are morally equivalent and do not think that it is at odds with core progressive values to have particular measures for certain general situations in a society.

It is my impression, and I have said this before, that for some reason I can guess at, in America it is more often the case that issues are viewed and analysed from the (arguably sometimes narrow) point of view of individual members of society. They may also, I'm not sure, in a possibly related way, tend to be somewhat short-termist.

 

[Don2 (Don1 Revised)]

I can't think of a better recruiting tool for the KKK.

Really? I heard they have good retirement plan benefits.

How about addressing the point?

What point: you used hyperbole? People are racist and they will use information to be racist and as a recruiting tool. If there is no valid information, then they will make stuff up. Conspiracies, whatever. Therefore, the way we do things should not be impacted much by how racists will react. Instead, we should do things because they are the right thing to do.

 

[Metaphor]

For all the people who think being a particular race can be "performance enhancing" merely due to being that race (e.g. because they are more trusted by the public):

Would you make the same argument about female surgeons? There is evidence that men and women have a preference for male surgeons; they are seen as more trustworthy and competent. Should colleges of surgery preferentially accept male students on that basis? Why or why not?

If you don't think they should have a preference system, why should the prejudice of the public play a role in who gets recruited into the public service?

 

[ruby sparks]

Plus, most of the people in the majority group negatively impacted by such policies will be those who are below average in advantages overall, b/c they are the one's closer to the cutoffs who are the first to cut rejected when being in that group is weighted negatively. Bear in mind that C thru F logically entail using different cutoff criteria for people of different groups. Mathematically, that is what using group membership as a weight in the algorithm means. It lowers the criteria needed for the minority group and raises the criteria for the majority group. So, it is those near the cutoff who get cut when their criteria for their group is raised. So, even people who don't have a principled moral concern about individual-devaluing discrimination, have a reasonable and rational basis to object to policies most likely to cause them or their loved one's harm. Plus, those most harmed are those with the most in common with the minority group in terms of SES, and thus this resentment undermines what should be a socio-political alliance.

Sorry, I forgot to deal with this one.

Sure, that is another reason for backlash. You have thus offered two. The first, the one based on moral principles, I have already commented on. The one above is, again, valid, but I wonder about its applicability (how many are actually harmed, how great is the harm, etc) and whether it is, in reality, 'reasonable and rational'. For example, how come, in 2014, a majority of white Americans polled could agree that, “today discrimination against whites has become as big a problem as discrimination against blacks and other minorities”?

And what about the other one, that comes from the same source as the unfair discriminations which are at least part of the reason remedies are even considered. Plain, old-fashioned racial resentment. Don't leave that one out. It's apparently quite pervasive. Exactly why is a very interesting question. I suspect it's partly natural, human ingroup/outgroup behaviour, and partly because it's stoked up by those with power and influence, who play heavily on the fears of those who listen to them.

So in that sense I do agree with your general point about backlash (rational, racist or otherwise) preventing the sort of socio-political alliance which might address what might arguably be called the 'even bigger issue'' and maybe even at a pinch 'the real issue'. It doesn't take much imagination to think that preventing such an alliance is part of the strategy for those at the top of the wealth and power pyramid in the USA, and possibly has been for a very long time. Divide and conquer and all that. Or to put it another way, Donald Trump, if he is a racist, is most likely much more interested in the colour of money than the colour of skin.

Which I think, in America, may make him.....a fan of the Greens, of sorts.

 

[Loren Pechtel]

How about addressing the point?

What point: you used hyperbole? People are racist and they will use information to be racist and as a recruiting tool. If there is no valid information, then they will make stuff up. Conspiracies, whatever. Therefore, the way we do things should not be impacted much by how racists will react. Instead, we should do things because they are the right thing to do.

The issue is when you punish people for being white you'll get backlash.

 

[Trausti]

How about addressing the point?

What point: you used hyperbole? People are racist and they will use information to be racist and as a recruiting tool. If there is no valid information, then they will make stuff up. Conspiracies, whatever. Therefore, the way we do things should not be impacted much by how racists will react. Instead, we should do things because they are the right thing to do.

The issue is when you punish people for being white you'll get backlash.

Hey, when your at the bottom of the intersectional totem pole you gotta learn to kiss the ring.

 

[ruby sparks]

I'm going to throw a few very interesting things into the hypothetical mix.

The first (which is actually a change to the original OP scenario) is that there aren't just or only greens and blues on the planet. There are a number of other, smaller minorities, the main two of which are pinks and oranges, though there are others too.

The second (which isn't a change to the OP scenario) is that the authorities on the planet have been very relaxed about intermarriages between the different races, so as a result there is also a non-insignificant (and growing) minority of greeny-blues, bluey-greens, greeny-pinks, orangey-blues and so on.

The third (which again is not a change to the OP scenario, just additional to it) is that there are numerous neighbouring planets, some of which are majority green and some of which aren't. Travel between the planets is relatively simple, which allows for immigration and emigration, albeit in the case of our (the OP scenario) planet, there is net immigration pressure (it's an attractive planet for a variety of reasons) and also there are more non-greens trying to migrate there than there are greens.

In fact, various demographic trends (including slow population growth and an ageing profile for greens, coupled with faster growth and a more youthful demographic profile for all the non-greens) mean that it is already the case, for example, that in the under-5 age category, there is already on the planet a majority of non-greens. Also, by next year, the same will be true of under-18's. Demographic projections further ahead predict that in 25 years, greens will be in the overall minority (below 50% of the population) on the planet, and this trend is expected to continue (they are predicted to be only a third of the total population in a mere 40 years time) albeit they may still be the largest single group.

Such a set of fluid circumstances may be seen perhaps to weaken the case slightly for implementing AA measures of any sort (A through F) and perhaps E & F especially, because what is going to happen is that population change and pressure alone may ensure increasing racial diversity. Green critics of AA on the planet say that the problem is already, albeit slowly and gradually, solving itself, to at least some extent. Critics of those critics question this, and thus the debate over whether AA is (a) still necessary, (b) racism and/or (c) a good or bad thing, continues, but even AA supporters acknowledge that the phenomena described in this post subtly alter the dynamics of the situation. 'True progressives', even among the generally more privileged greens, welcome the trend, while other greens greatly fear for the future of greens, a reaction which can be partly rational and partly irrational (or because they are non-progressive and/or want to hold on to their privileges).

 

[Don2 (Don1 Revised)]

How about addressing the point?

What point: you used hyperbole? People are racist and they will use information to be racist and as a recruiting tool. If there is no valid information, then they will make stuff up. Conspiracies, whatever. Therefore, the way we do things should not be impacted much by how racists will react. Instead, we should do things because they are the right thing to do.

The issue is when you punish people for being white you'll get backlash.

And the backlash you should be concerned about, if at all, is from good people, not racists. Moreover, (and again), you ought to do it or not do it because it's right or wrong. Backlash is less of a consideration these days since no matter what, you get backlash.

 

[ruby sparks]

For all the people who think being a particular race can be "performance enhancing" merely due to being that race (e.g. because they are more trusted by the public):

Would you make the same argument about female surgeons? There is evidence that men and women have a preference for male surgeons; they are seen as more trustworthy and competent. Should colleges of surgery preferentially accept male students on that basis? Why or why not?

If you don't think they should have a preference system, why should the prejudice of the public play a role in who gets recruited into the public service?

Unless there are exceptional or discretionary circumstances or other potential practical benefits which may outweigh the downsides, prejudices, and other non-rational (or non-justified) perceptions and preferences, should not, imo, be the reason for deciding. Let's just say that by and large they should not.

On the other hand, if there are reasonable grounds to argue that, for example in the police, there has been and is unfair discrimination by the (mainly green) force, then the preferences are not mere prejudice and can be somewhat justified. Ditto if there has been actual, similar discrimination happening in or by, say, other public or governmental authorities or departments.

To again briefly stray from the OP scenario, the arguably slightly different situation regarding, say, doctors, can be complicated by the fact that as already mentioned, green doctors are more reluctant than non-greens to work in mainly non-green (usually poorer) locations, which can lead to shortages of doctors in those places.