Affirmative Action

[Deviate]

This is my first attempt at transferring a thread for the Tavern to the Chamber so we'll see how this goes...

http://networkedblogs.com/xFddK

Elizabeth Warren... running for office. Some big scandal has broken out because she claims Native American heritage. There are questions about the truth in this, troublesome documents from over a hundred years ago, etc.

She's used a claim of 1/32 NA, and this helped her get her job as a professor at Harvard.

A series of questions...
1) What discrimination, etc has Warren faced due to her 1/32 NA heritage?
2) Does anyone see any discernable traits of a NA in her?

Spoiler :

3) At what point do we say, ok, you can't claim that? 1/64th? 1/128th?

The ultimate irony, she could have used her claim to edge out someone who is 100% NA or whatever quota might have needed to be filled at the time.

Why the hell are we giving people jobs based on race?! It shouldn't be a consideration at all, is my point of view... now we draw completely arbitrary lines in the sand, and good people get screwed over by people doing stuff like this.

 

[Deviate]

I'm obviously cutting out a lot of the conversation that has gone on in the ~13 pages in the Tavern but hopefully we can pick up here...

I see racial quotas in the same way that I see minimum wage laws. A minimum wage is effectively a quota on the amount of compensation an institution must legally be required to meet, and the academic literature is clear that both quotas and minimum wages are terrible ideas. The minimum wage is clearly economically stupid; ceteris paribus it will drive up costs, result in greater unemployment, feed inflation, and result in a misallocation of capital. It's unfair to burden employers and similar institutions with the task of reducing inequality - that should be the government's job. There are other ways of reducing inequality, and we already do those, so why do we need the minimum wage? SiLL's arguments parallel these arguments remarkably closely -- and, like the anti-minimum wage arguments, they miss the point entirely.

We know that there are bad effects of the minimum wage, just like we know that there are bad effects of AA. We hear anecdotes that employers can't hire a white guy because they can't afford to pay him minimum wage; we hear anecdotes that employers can't hire the marginally better qualified white guy because the black guy got +5 bonus points due to his race. We know that it can cause injustices here and there. We know that it isn't particularly efficient. We know that it is a blunt tool that has numerous problems. But we also know that it gets the job done. We know that over the long term, such laws rectify social problems that we simply could not solve through any other means. We know that driving up wages artificially actually creates positive feedback in the long-term. And we know that putting members of a disadvantaged group in artificially higher positions, even though there is a marginally more qualified Straight White Male available, creates a positive feedback among that group. When a black guy or a woman or a homosexual or a man who grew up in crippling poverty in an inner city look at universities, factories, offices, etc and see people like them there, they think that they can do it themselves too, and it becomes self-fulfilling.

For me, wage quotas and points-based affirmative action are two sides of the same coin. They're blunt tools with well publicised ill-effects. But you know what? They work. We're artificially rectifying a problem in a few generations that would otherwise take centuries to solve, if at all.

 

[illram]

Moderator Action: Excellent idea. This will be a case study on how this forum is different from the Tavern however. The tone in the quoted OP, for example, won't fly here, whereas the quote in the above post is a quality example.

 

[luiz]

If it's true that Elizabeth Warren had any advantages because of having (supposedly) 1/32 of indian ancestry it'll become one of my favorite examples of the absurdity of racial policies (along with the identical twins who were classified in different races by a Brazilian university).

@Mise: who is "we"? I certainly don't agree with the benefits of racial quotas. I think that decades of race-based AA worldwide, from Malaysia to the US, have produced far more harm than good. The minorities who catched up with the majority group didn't need AA; the minorities who get AA never caught up. It's an abject failure.

 

[Terxpahseyton]

I'll follow the call to the Chamber and will add my exchange with Mise so far beginning with the quote of the OP.
I'll for context also add most of my response to Masada from which I reused a part in my conversation with Mise.

For me, wage quotas and points-based affirmative action are two sides of the same coin.

For me they are not, because while I have nothing to argue against your comparison, the point of minimum wage is to actually help people in a disadvantage, while the point of race-based AA is to help races in disadvantage.
And to quote myself

To look at the statistics, to see there the statistical disadvantage of black people and to then conclude that black people deserve help, that's cool with me. That this is supposed to trump the fact that people in general need help is not. I see only one justification for such an attitude: That it is more important to support disadvantaged races rather than individuals. And this is, while maybe not racist in the original sense of the word, still race-inspired pure awfulness if you ask me. It is still a thinking that on all levels weights the interests of one race higher than the other, just in the misguided opinion that this was justified. But I got to ask directly: If we have a disadvantaged black kid and a disadvantaged white kid, which while for different reasons hold the same socio-economic disadvantage, does the white kid deserve less support because his ancestors were not enslaved?

(Almost) full post:

Spoiler :

I am honestly not sure what your point is. Oppressed minorities can (and usually do) to this day inherit the woes suffered by their ancestors, be it African-Americans or Australian indigenous. Which on average will put members of those minorities in a worse position than the national average. I understand this and am actually surprised that you feel the need to go into such lengths to describe this.
I in deed also support the wish to fight this. However, I don't think that a member of such a minority which is not well off due to socio-economic reasons deserves to be helped any more than a member of the majority which isn't well off due to socio-economic reasons. Hence my suggestion to focus on the socio-economic background. That starts with family income, it can go on with the parents educations, a bad neighborhood where someone grew up or whatever. I haven't exactly cracked my head over all the possibilities, but it is my impression that a lot of factors can be used which are - on the indivdual level - more just than something as encompassing and general as race.

To look at the statistics, to see there the statistical disadvantage of black people and to then conclude that black people deserve help, that's cool with me. That this is supposed to trump the fact that people in general need help is not. I see only one justification for such an attitude: That it is more important to support disadvantaged races rather than individuals. And this is, while maybe not racist in the original sense of the word, still race-inspired pure awfulness if you ask me. It is still a thinking that on all levels weights the interests of one race higher than the other, just in the misguided opinion that this was justified. But I got to ask directly: If we have a disadvantaged black kid and a disadvantaged white kid, which while for different reasons hold the same socio-economic disadvantage, does the white kid deserve less support because his ancestors were not enslaved?

People do respond, that white people simply don't have as much trouble. Well sure, that is why looking at the socio-economic background will likely result in a majority of non-whites profiting. And that then is perfectly fine and just.

But as it is, AA seems to create a ton of situations where there is no justice to speak of. Where minorities simply take advantage of this program to unduly profit from it. And of course they do and can, because nobody even asks otherwise.

Which means, that my beef is in the end not primarily with the practical shortcomings of race-based AA (though they are part of the argument), like it were in case of someone arguing against minimum wage, but with the point of such a measure to begin with. Which is to help races instead of individuals. If such a race really is in need, helping individuals in need should totally suffice.

-------------------------------------------------------------------------------

Mise, I am afraid I still don't agree with you, but I will make an effort to make everything as comprehensive and clear as necessary. I would appreciate it if you respected that and make room for the possibility that I may too have a point worth of consideration.

No, he doesn't "deserve" less support. Everyone is deserving of however much support they need to live a fulfilling life; as a Liberal that is something I believe in passionately and will defend vehemently. And, exactly to that point, the white kid doesn't need as much support, because the challenges facing an under-privilaged white man are significantly less difficult to overcome than the challenges facing an under-privilaged black man. In addition to being underprivilaged, the black man is also black, and faces problems that a white man will simply never face. As I said in response to kochmann, we know that when you control for all possible factors, including socio-economic background, blacks, women, gays etc are still disadvantaged; they get paid less, they are given fewer interviews, they are given fewer university places, they are given fewer opportunities at work, fewer opportunities for promotion, and so on. That's just a fact, and I implore you to seriously consider that your assumptions are just flat out wrong. You can't say that socio-economic factors are the only important factors; race, gender, sexual orientation, disabilities, and so on -- these are all factors as well by themselves. You, in particular, much more than certain others in this thread, are intellectually capable enough, and intellectually curious enough, to both explore and understand the quantitative statistics and qualitative problems that disadvantaged groups face on a daily basis. There is a reason why certain groups are labelled as being "disadvantaged", and it's not because we on the liberal left have all suddenly succomb to collective lunacy and turned into dispicable racists.

I know that this is commonly accepted as a fact, and I agree, but do you know actually of a single study that does control for socio-economic and at best also cultural factors and which shows the actual significance of race for success in life? I just looked on google for half an hour and did not find a thing (I could tell you though that even though black women die more often from breast cancer, race does not actually play a significant role ). To know the actual dimensions would probably be quit helpful, but I don't know where to take it from.

Anyway, let me expand on what you said. You differentiated between racial discrimination and socio-economic factors. Which is important for my argument as well.
I talked a lot about how socio-economic factors were superior to racial factors and you respond by claiming that socio-economic factors would not account for contemporary racial discrimination. This is wrong. It does indirectly, because any socio-economic situation will also be a result of race, as I am sure you will agree. Additionally, it in this sense will also ensure that those members of a given race really benefit, which are also in actual apparent need to do so.

Now you say racial AA is justified in that it does further control for such discrimination. And that of course is true. I don't doubt that this has an impact on how well a race does. Now the obvious drawback of this approach is that it is every unspecific. In practice, it will assume every member of a racial minority to be discriminated against. And that plain and simple means that racial AA is not a measure to duly benefit disadvantaged individuals, but as said a disadvantaged race. On an individual basis, as I already said in another post, IMO something resembling justice can not be expected and quit ordinarily, the opposite will take place. For the natural reason, that race is, on the individual level, not a statistic, but a very diverse influence, which will in many cases not mean a significant disadvantage. And that, I would say, is also a fact.

And given that already socio-economic approaches will benefit minority races dis-proportionally and hence will do a lot in leveling the playing field, I am left to wonder why such a crude instrument as AA is still seen as necessary. And I suggest, that that is so because people are more interested in improving the statistical success of a race than individuals which are actually in need. That people are so focused on those statistics, that they loose the actual reality -taking place on individual levels - out of sight.

 

[luiz]

I know that this is commonly accepted as a fact, and I agree, but do you know actually of a single study that does control for socio-economic and at best also cultural factors and which shows the actual significance of race for success in life?

Actually, I've done some econometric studies using American data myself, and once you control for socio-economic factors (from parents education level to growing up on a broken home), you find that race does not usually play a statistically significant role.

 

[Terxpahseyton]

That sounds interesting, but can you expand on that or show data or whatever? I am sure you can imagine that it is hard to just take your word for it.

 

[JollyRoger]

If it's true that Elizabeth Warren had any advantages because of having (supposedly) 1/32 of indian ancestry it'll become one of my favorite examples of the absurdity of racial policies (along with the identical twins who were classified in different races by a Brazilian university).

What would you think of a 1/32nd blood becoming Chief of a tribe?

 

[Mise]

@Mise: who is "we"? I certainly don't agree with the benefits of racial quotas. I think that decades of race-based AA worldwide, from Malaysia to the US, have produced far more harm than good. The minorities who catched up with the majority group didn't need AA; the minorities who get AA never caught up. It's an abject failure.

The "we" I'm talking about is merely "people who agree with AA". Or perhaps liberal left-wingers in general, since the arguments in favour of AA typically coincide with arguments in favour of anti-discrimination laws more generally, and they are typically espoused by the liberal left.

-------

@SiLL, I'm going to answer in a different order to the way you posted, because I think it will help to structure the discussion more clearly.

Anyway, let me expand on what you said. You differentiated between racial discrimination and socio-economic factors. Which is important for my argument as well.
I talked a lot about how socio-economic factors were superior to racial factors and you respond by claiming that socio-economic factors would not account for contemporary racial discrimination. This is wrong. It does indirectly, because any socio-economic situation will also be a result of race, as I am sure you will agree. Additionally, it in this sense will also ensure that those members of a given race really benefit, which are also in actual apparent need to do so.

(highlighting mine)

Sorry, I didn't understand the last sentence, so I hope I'm not misinterpreting this part of your post. You say "this is wrong" - is that not what we are trying to establish? Isn't it absolutely essential that we reach an agreement on whether racial discrimination, even when you control for socio-economic factors, still exists?

To put it more clearly, if I can prove to you that, when you control for all socio-economic factors, a black man will still have fewer opportunities, lower wages, and so on, than an equivalently poor white man, will you agree that AA is necessary?

The rest of your post seems to rest on the assumption that socio-economic factors are sufficiently correlated with race to eliminate all differences between races. I agree that, if there was no racial discrimination in the West, and if all differences between races could be explained by socio-economic factors, that tackling tackling poverty etc would be enough to wipe out racial differences. What I'm saying is that this assumption isn't true, and various studies have shown this, so I hope we can narrow the discussion to just this one point: Can differences between races be explained entirely by socio-economic factors?

Mise, I am afraid I still don't agree with you, but I will make an effort to make everything as comprehensive and clear as necessary. I would appreciate it if you respected that and make room for the possibility that I may too have a point worth of consideration.

I know that this is commonly accepted as a fact, and I agree, but do you know actually of a single study that does control for socio-economic and at best also cultural factors and which shows the actual significance of race for success in life? I just looked on google for half an hour and did not find a thing (I could tell you though that even though black women die more often from breast cancer, race does not actually play a significant role ). To know the actual dimensions would probably be quit helpful, but I don't know where to take it from.

Well, let's take the example Deviate posted in the Tavern thread: http://www.nber.org/papers/w9873.pdf

In this case, they have strictly controlled for every variable. They are sending out 4 CVs to each company, randomly selected from 4 different "buckets" of CVs. The CVs get assigned random names, depending on whether they are in one of the "black" buckets or one of the "white" buckets, so the CVs themselves are necessarily and strictly uncorrelated with any other variable that is present on the CV. The only difference is the name*. The good thing about CVs is that it contains everything that is necessary to judge employment; put another way, a CV contains all possible factors that could influence an employers' willingness to give the candidate an interview. So we know for a fact that, once you randomise the names of the CVs, all other possible callback-influencing factors are controlled for completely.

Now, your objection in the Tavern thread, that employers are using "black names" as a proxy for socio-economic status (or education or work experience etc), is also controlled for: employers know the socio-economic status of the candidate, because it says so on the CV. Indeed, the researchers attempted to use the data to determine the level of "affluence-based" discrimination: employers judge candidates who come from affluent areas to be more employable (and therefore give more callbacks) than candidates who come from less affluent areas. And yet, the racial discrimination is still present: blacks still get fewer callbacks from whites, even when they both live in affluent neighbourhoods.

This study, incidentally, should be totally conclusive to you. All other variables are controlled for strictly; the CVs are randomly assigned "black" or "white" names, and since all other data relevant to employment is listed, there is nothing else to consider. The differences, which are statistically significant, show a marked difference between black and white sounding names. There's really no room for negotiation here - it was a remarkably well-designed experiment, that avoids the common pitfalls of other experiments (most notably, exactly what variables they should control for).

Spoiler *- :

*-The other difference is that some CVs are given more experience, more education, etc than others, labelled as "High quality" CVs in the paper. So really the 4 buckets can be broken down as follows:

A1) White name + High quality CV
A2) White name + Low quality CV
B1) Black name + High quality CV
B2) Black name + Low quality CV

We would just be interested in A vs B. The other relationships, for example the callback rate between B1 and B2 vs A1 and A2 etc are also interesting, but they aren't really important for this discussion. We're really interested purely in the difference between buckets A and B.

 

[luiz]

That sounds interesting, but can you expand on that or show data or whatever? I am sure you can imagine that it is hard to just take your word for it.

Yeah, I'll try to come up with the studies I did myself.

In the meantime, you can Google the numerous writings on the topic by Thomas Sowell. He was the one to first demonstrate that things like growing up in a broken family matter far more than race, in fact making race statistically insignificant.

Well, let's take the example Deviate posted in the Tavern thread: http://www.nber.org/papers/w9873.pdf

In this case, they have strictly controlled for every variable. They are sending out 4 CVs to each company, randomly selected from 4 different "buckets" of CVs. The CVs get assigned random names, depending on whether they are in one of the "black" buckets or one of the "white" buckets, so the CVs themselves are necessarily and strictly uncorrelated with any other variable that is present on the CV. The only difference is the name*. The good thing about CVs is that it contains everything that is necessary to judge employment; put another way, a CV contains all possible factors that could influence an employers' willingness to give the candidate an interview. So we know for a fact that, once you randomise the names of the CVs, all other possible callback-influencing factors are controlled for completely.

Now, your objection in the Tavern thread, that employers are using "black names" as a proxy for socio-economic status (or education or work experience etc), is also controlled for: employers know the socio-economic status of the candidate, because it says so on the CV. Indeed, the researchers attempted to use the data to determine the level of "affluence-based" discrimination: employers judge candidates who come from affluent areas to be more employable (and therefore give more callbacks) than candidates who come from less affluent areas. And yet, the racial discrimination is still present: blacks still get fewer callbacks from whites, even when they both live in affluent neighbourhoods.

This study, incidentally, should be totally conclusive to you. All other variables are controlled for strictly; the CVs are randomly assigned "black" or "white" names, and since all other data relevant to employment is listed, there is nothing else to consider. The differences, which are statistically significant, show a marked difference between black and white sounding names. There's really no room for negotiation here - it was a remarkably well-designed experiment, that avoids the common pitfalls of other experiments (most notably, exactly what variables they should control for).

Spoiler *- :

*-The other difference is that some CVs are given more experience, more education, etc than others, labelled as "High quality" CVs in the paper. So really the 4 buckets can be broken down as follows:

A1) White name + High quality CV
A2) White name + Low quality CV
B1) Black name + High quality CV
B2) Black name + Low quality CV

We would just be interested in A vs B. The other relationships, for example the callback rate between B1 and B2 vs A1 and A2 etc are also interesting, but they aren't really important for this discussion. We're really interested purely in the difference between buckets A and B.

Well, I have some objections to the study. They're comparing "normal white" names, like Emily and Greg, to poor, uneducated black names like Lakisha and Jamal. The paper's title uses those examples.

If you look at table 11, with all the names used, you'll see all "white names" are fairly normal while all "black names" are fairly bizarre. There was no "poor white" bizarre name used, like Cletus or Billy Bob. The fact that they failed to test the response to "poor white" names seems a bit damning to me. Their method to choose names, the "most uniquely white" and "most uniquely black" ones, is good to convey the race of the applicant to the employer, but fails at testing alternative hypothesis for discrimination. Also the fact that they only looked at babies born in Massachusetts and Chicago is a potential methodological flaw, as white poverty is mostly a rural phenomena and as such poor white names won't be found there in significant numbers, while poor black names will (black poverty is overwhelmingly urban). Would the study reach the same conclusions in Appalachia?

Fact is even if employers know the socio-economic background of the applicants, they can still infer stuff from the names, and I'm not talking race. If you take a Cletus and a Brad, both white and both with the same family income, I'm willing to bet the Brad will receive a better response to his sent CVs. To me it's fairly obvious that certain white names would also face discrimination.

Another weakness of their name-choosing method, which they admitted themselves, is that it does not represent the average names of Black Americans, who in general have much "less distinctive" names. So their names represent a particular sub-set of Black Americans, and that subset (a rather poor, uneducated subset) certainly faces discrimination also from non-racial factors. In other words, the conclusions cannot be extrapolated to all Black Americans.

At any rate, I would much prefer broader economic studies about salaries and credit access, using much larger data bases, to this rather limited and necessarily subjective social study.

So, was the study conclusive? I don't think so.

 

[Mise]

Again, as I said, they specifically looked at the socio-economic status of the candidates as well. I'll just quote what they say about it:

5.1 Potential Confounds

Though we have interpreted our results in terms of racial differences, we actually manipulate only the
name on the resume. While these names clearly signal race, perhaps they also signal some other personal
characteristics. More specifically, one might be concerned that employers are inferring social background
from the personal name. When employers read a name like “Tyrone” or “Latoya,” they may assume that the
person comes from a disadvantaged background. In the extreme form of this social background interpretation,
employers do not care at all about race but are discriminating only against the social background conveyed
by the names we have chosen.
47
While plausible, we feel that some of our earlier results are hard to reconcile with this interpretation.
For example, in Table 6, we found that while employers value “better” addresses, African Americans are
not helped more than Whites by living in Whiter or more educated neighborhoods. If the African American
names mainly signal negative social background, one might have expected the estimated name-gap to be
lower for the better addresses. Also, if the names mainly signal social background, one might have expected
the name gap to be higher for jobs that rely more on soft skills or require more inter-personal interactions.
We found no such evidence in Tables 6 or 7.

(best viewed in original pdf format!)

They go on to acknowledge that the "black" names they chose were associated with below-average socio-economic backgrounds. However, they use their data to directly test whether names associated with below-average socio-economic backgrounds are correlated with low callback rates. They find no evidence of this; in fact, they find that the exact opposite is true, especially in the African Male category (which is the only statistically significant result):

But, more interestingly for us, there is substantial between-name heterogeneity in social background.
African American babies named Kenya or Jamal are affiliated with much higher mothers’ education than
African American babies named Latonya or Leroy. Conversely, White babies named Carrie or Neil have lower
social background than those named Emily or Geoffrey. This allows for a direct test of the social background
hypothesis within our sample: are names associated with a worse social background discriminated against
more? In the last row in each gender-race group, we report the rank-order correlation between callback
rates and mother’s education. The social background hypothesis predicts a positive correlation. Yet, for all
four categories, we find the exact opposite. The p-values indicate that we cannot reject independence at
standard significance levels except in the case of African American males where we can almost reject it at
the 10 percent level. In summary, this test suggests little evidence that social background drives the extent
of discrimination.

In other words:

If you take a Cletus and a Brad, both white and both with the same family income, I'm willing to bet the Brad will receive a better response to his sent CVs.

You would have lost this bet!


I find it very hard to criticise this paper in particular, which is why I'm keen to focus on it p). Economic research by the NBER is typically very rigorously conducted and tested, and this is no exception.

 

[luiz]

What would you think of a 1/32nd blood becoming Chief of a tribe?

If membership of the tribe is open to anyone who feels culturally associated to the tribe, I'd praise them.

If they require a certain "blood quota", be it 1/32 or 1/3200, I say damn them.

 

[luiz]

Again, as I said, they specifically looked at the socio-economic status of the candidates as well. I'll just quote what they say about it:

(best viewed in original pdf format!)

They go on to acknowledge that the "black" names they chose were associated with below-average socio-economic backgrounds. However, they use their data to directly test whether names associated with below-average socio-economic backgrounds are correlated with low callback rates. They find no evidence of this; in fact, they find that the exact opposite is true, especially in the African Male category (which is the only statistically significant result):




I find it very hard to criticise this paper in particular, which is why I'm keen to focus on it p). Economic research by the NBER is typically very rigorously conducted and tested, and this is no exception.

I'm finding stuff to criticize it. First, as I said, they are only testing their hypothesis for a particular subset of blacks, which does not represent the average at all. I don't see how those results can be freely extrapolated to all blacks.

Also, the way they tested the social hypothesis seems deeply flawed to me. While it may be true (if they so it surely is) that "Kenya or Jamal are affiliated with much higher mothers’ education than African American babies named Latonya or Leroy", that does not at all translate to popular perception! I don't think people differentiate at all between Jamal and Leroy, they're both "poor black names" as far as employers see it. They should have tested a black "Daniel" or "John" versus a black "Leroy" or "Tyrone", making sure the employers know their ethnicity. Or a white "Brad" vs a white "Cletus", as I suggested. Their conclusion that there is no discrimination based on signaling of social background by names is just wrong. I see people discriminated by names all the time (with no racial undertones), and I'm sure it happens everywhere. The fact of the matter is they did not test the social hypothesis at all.

So I would not have lost the bet, because that hypothesis was not tested. They tested a certain "less poor" black name vs a "more poor", assuming employers recognize the difference, which they most certainly don't.

 

[Mise]

I'm finding stuff to criticize it. First, as I said, they are only testing their hypothesis for a particular subset of blacks, which does not represent the average at all. I don't see how those results can be freely extrapolated to all blacks.

Also, the way they tested the social hypothesis seems deeply flawed to me. While it may be true (if they so it surely is) that "Kenya or Jamal are affiliated with much higher mothers’ education than African American babies named Latonya or Leroy", that does not at all translate to popular perception! I don't think people differentiate at all between Jamal and Leroy, they're both "poor black names" as far as employers see it. They should have tested a black "Daniel" or "John" versus a black "Leroy" or "Tyrone", making sure the employers know their ethnicity. Or a white "Brad" vs a white "Cletus", as I suggested. Their conclusion that there is no discrimination based on signaling of social background by names is just wrong. I see people discriminated by names all the time (with no racial undertones), and I'm sure it happens everywhere. The fact of the matter is they did not test the social hypothesis at all.

So I would not have lost the bet, because that hypothesis was not tested. They tested a certain "less poor" black name vs a "more poor", assuming employers recognize the difference, which they most certainly don't.

Well, I couldn't disagree more... Even if there are differences between popular perception of the socioeconomic status of certain names and the reality of what those names actually correlate with, the fact that they find the exact opposite of what your theory suggests is pretty damning to your theory, IMO...

In any case, the fact that, on a black candidate's application, the only thing that matters was his name* is surely evidence that people are so god damn awful at judging the ability of a candidate that preserving the process of judging the ability of a candidate by his CV is of questionable value in the first place. If the only thing that matters on a CV with the name "Jamal" or "Cletus" on the top is his name, then surely, following SiLL's original logic, we should be assigning quotas to companies based on the name "Jamal", "Cletus", etc, rather than race. Would you accept that as a solution? I don't think you would, but that's the conclusion you're drawing, isn't it?


*-Table 5. Actually, "special skills" was also significant, but only by a tiny amount.

 

[luiz]

But the thing is, they did not test the hypothesis that a black with a "normal" name would do better than a black with an "uniquely black name" (in reality, a poor name). What they did was to select some of the "uniquely black names" that correlated with better socio-economic status and tested against some that correlated with worse. That says nothing of my hypothesis, it wasn't tested at all. The result of their test was almost certainly spurious.

Here's what I would like to see tested, but wasn't:

-A poor white name versus a "normal white name": say, Cletus vs. Brad.
-A poor name versus a "normal name", while informing that both applicants are black: say, Jamal vs. John.

I'll insist once again: this study was only about a certain sub-set of the black population, which does not represent the average at all, and I don't see how those results can be freely extrapolated to all blacks.

And no, my solution would not be adopting quotas for people with poor names. I don't think we should adopt quotas at all; if we had to adopt them for all characteristics that correlate negativity with job prospects, we would have to adopt quotas for short people, ugly people, etc.

It's a fact that people discriminate on names all the time, and I think parents should think of that.

I don't have (and don't think we should pursue) any solution (other than parents thinking for a couple of minutes before choosing a name). I abhor the notion that we should create policies to counteract every social correlation we observe.

 

[Mise]

Yes, I understand that that's what you want tested, and I understand that this wasn't tested. As you say, the "black" names are biased toward the low end of the socioeconomic spectrum. But what are you actually saying here? That employers are thinking:

1) "Jamal" is on the low end of the socio-economic spectrum
2) People on the low end of the socio-economic spectrum usually have poor qualifications and weak employment histories
3) I don't want someone with poor qualifications and weak a employment history to work for me
4) I won't hire anyone named "Jamal"

But the employer already has Jamal's CV right in front of him, staring him in the face. To see his qualifications, he has to look 10cm down the page. To see his employment history, he has to look 20cm down the page. Hell, to see his socioeconomic background, they only have to look at the 2nd line underneath his name: his address! Are you trying to tell me that employers, as a group, are so stupid and lazy that they won't even read a candidate's CV to find out what their actual education or work experience is? That they would, instead, make a series of assumptions and leaps of faith, rather than glancing down a few inches to find out the facts?

If that's what you're saying, then why on earth are you so desperate to preserve this crazy, illogical and economically inefficient method of allocating labour? Surely we need a better way. Perhaps points-based AA is merely ironing out the inefficiencies in the way we allocate labour.

And no, I don't expect you to agree with this solution. But if you want me to believe that employers completely ignore what's written in front of them in Jamal's actual CV and instead concoct a fictional, stereotypical view of what someone with the same socioeconomic background as "Jamal" may hypothetically have on his CV in their own heads, then I do expect you to come up with some evidence for it. Because I simply don't have such a dim view of business owners as you In short, I don't find your theory, that employers aren't racist, but merely stupid, lazy, and prejudiced in a slightly less reprehensible way, as evidence that we should ditch AA. On the contrary, we need to ramp it up!

 

[Terxpahseyton]

I'll post a thorough reply later, but I wanted to say that it is perfectly normal that names which sound poor are discriminated against for no other reason than that they sound poor. You'll have this in any other country of the world. This may sound like a thing too irrational to be true, but that's how people work. You don't actually need a conscious ill intend for that. To illustrate: Beautiful women are similarly discriminated against because women often take care of the admission process and will instinctively tend to rule out beautiful women because they instinctively perceive them as rivals. While beautiful men score.
Likewise, a poor name will at least on the level of the sub-conscious trigger negative impressions. And that is what exotic black people names appear to be: poor people names.
Society is full such dynamics and if we want know how it really matters that someone is black, we need to account for them.

 

[Mise]

So there is discrimination based on:
- attractiveness
- "poor-sounding" names
- height
- gender
- baldness
- accents
- acne
- speech impediments
- clothing
- sports ability
- musical ability

But there cannot possibly be discrimination based on:
- race

 

[Terxpahseyton]

Jesus Christ, what the hell is this?!

This was about the merits of this particular study and nothing more. I hate to pull it, but straw man much? Can you not just accept that your line of thought was flawed, do you need to move to such a dubious attack right ahead? I know that this kind of debating style may be entertaining or whatever, but it is also freaking indecent and unproductive.

 

[Mise]

Err, maybe you should calm down...

You were saying that the discrimination could be a result of poor people's names being used, rather than black people's names being used. I'm asking why you seem to believe that. I asked luiz for evidence that employers actively discriminate against poor people; you respond by saying "it is perfectly normal that names which sound poor are discriminated against for no other reason than that they sound poor". Well, why can't I just say "it is perfectly normal that names which sound black are discriminated against for no other reason than that they sound black"? That's not evidence, that's merely a restatement of your assumption that poor-sounding names are more important than black-sounding names in the employers' discrimination process. You seem so quick to conclude that employers can be prejudiced against pretty girls, and that this is a crucial confounding variable. But you and luiz both seem unwilling to accept that employers can be prejudiced against black people at all, since you believe that any discrimination based on race can be ruled out by looking at socioeconomic variables instead.

I mean, your entire argument rests on the assumption that all differences in hiring decisions are due to the socioeconomic background of the candidate, and are not attributable to race. So what am I supposed to conclude? The only logical way that this could possibly be true is if employers never discriminated on race, but really discriminated on socioeconomic background instead. And, since this is what you asserted directly above, I don't see why you're getting upset about me repeating it in my own words.