Racism Proven Stupid by SCIENCE!

[Einhander Sn0m4n]

Slate: Amazing Racism

A study that shows the true stupidity of discrimination.
By Tim Harford
Posted Saturday, Feb. 2, 2008, at 7:39 AM ET

Hillary Clinton and Barack Obama notwithstanding, the world still seems to be ruled by white men. Is this the result of racial and sexual discrimination in the workplace? Or are other factors more important—for instance, that too few black kids attend good schools, or that women usually interrupt their careers to have children? The answer is far from academic, because if we want to change a situation, it's a good idea to work out what might be behind it.

Economists have been leading this investigation for longer than one might think. Contrary to popular belief, "the dismal science" did not acquire its name because of Thomas Malthus' gloomy predictions. The title was bestowed upon us in 1849 by Thomas Carlyle, who attacked John Stuart Mill and his fellow political economists for their "dismal" support for emancipation, and their insistence that former slaves, women, even the Irish, were all equal.

More recently, economists have turned their attention to the subtler question of how to detect discrimination in labor markets. That women or racial minorities are paid less suggests discrimination but does not prove it, even if the statisticians make every effort to adjust for other differences such as part-time work or different choices of jobs.

One solution to the ambiguity is a random audit. A recent example was carried out by economists Sendhil Mullainathan and Marianne Bertrand. They generated about 5,000 fake job applications and used a computer to add, at random, distinctively black or white names. The employers who received the applications systematically favored the Gregs and the Emilys over the LaTonyas and the Jamals. Perhaps even more perniciously, they paid attention to the qualifications of (apparently) white applicants, but did not notice the difference between mediocre black applicants and excellent ones.

But discrimination should also show up in another way. Employers who prefer not to employ workers because of their sex or the color of their skin are likely to lose money: Employing stupid white men when you could be employing smart black women is not a profitable human-resources policy. Employers might nevertheless do this, either because they do not realize that their prejudices are costing them money, or because they do not care. If so, discrimination is easy to detect in principle: Just note that the profitable firms will be the ones employing more women or workers from an ethnic minority.

Economist Stefan Szymanski realized that the English soccer league was a perfect testing ground for this hypothesis. There was excellent data available as to which clubs employed black players; soccer has a very clear measure of success, and unlike some sports leagues, the English game does not go in much for redistributing money from successful clubs to minnows.

Szymanski studied the game between 1978 and 1993, a time encapsulated by the image of Liverpool's Jamaican-born star, John Barnes, backheeling away a banana that had been hurled at him from the stands. But Szymanski's numbers suggest that it was the owners, not the fans, who were the worst offenders. Clubs that bucked the norm and fielded several black players did not suffer lower attendance or revenues as a result. But they did enjoy a higher league position with a lower wage bill than the typical club—clear evidence that black players were underpaid on racial grounds.

As soccer has become ever more competitive and the financial stakes have become higher, racial discrimination has become more expensive. I note that many more black players now play in the top division of British soccer. That is good news; it's also no coincidence.

If only competition was as fierce, and talent as undeniable, in the world outside the stadium gates.

[Enigma]

Though I agree that for a company to be profitable one must employ the best they can get but the article seems to say that white men = stupid while women and black people = very smart.

I understand the intent but not how it was worded.

[Alyrium Denryle]

Though I agree that for a company to be profitable one must employ the best they can get but the article seems to say that white men = stupid while women and black people = very smart.

I understand the intent but not how it was worded.

Only if you fail at reading comprehension.

If Company A passes over a skilled black worker for an unskilled white worker and does so consistently, it stands to reason that they will have lower profits.

[Einhander Sn0m4n]

Though I agree that for a company to be profitable one must employ the best they can get but the article seems to say that Iraqis = stupid while Republicans and Troops = very smart.

I understand the intent but not how it was worded.

Word Substitution FTW!

[Enigma]

Though I agree that for a company to be profitable one must employ the best they can get but the article seems to say that white men = stupid while women and black people = very smart.

I understand the intent but not how it was worded.

Only if you fail at reading comprehension.

If Company A passes over a skilled black worker for an unskilled white worker and does so consistently, it stands to reason that they will have lower profits.

Fuckhead I know that. But the article points out that women and black people would be better and would improve a companies profitability.

Just note that the profitable firms will be the ones employing more women or workers from an ethnic minority.

If you were paying attention to my reply, I agree that a company needs to hire the best. It shouldn't matter the race nor sex of the individual. If that person can better improve the company then hire him\her. That article doesn't go that route rather go about in a reverse discrimination.

[Enigma]

<snip>
Word Substitution FTW!

That was sooo good Einy that I concede all points. Congrats you are a very good debater. No, you are a master debator. All should bow down to you mad skills.

[Mayabird]

Not really science. Economics. Quite a different thing.

Also, this is VERY old news. I read it in a Milton Friedman book that was published decades ago. He was arguing that economics should end up fixing all the civil rights problems, because people would rationally see that discriminating against people means that they're held down too, and people who didn't discriminate would do better and win out in the end.

Of course, this ignores the fact that a good hunk of the world's population isn't rational in the least. Let's take the Deep South, for instance. Everybody knew it was the economic armpit of the country. But did they care? No, of course not. It was way more important to them to keep them darkies down than to, you know, be successful themselves.

[VT-16]

But the article points out that women and black people would be better and would improve a companies profitability.

No, it doesn't, moron. It uses a random situation as an example of how racism might cost a particular company. Read it again.

[Enigma]

But the article points out that women and black people would be better and would improve a companies profitability.

No, it doesn't, moron. It uses a random situation as an example of how racism might cost a particular company. Read it again.

I did but explain to me this quote.

If so, discrimination is easy to detect in principle: Just note that the profitable firms will be the ones employing more women or workers from an ethnic minority

I agree the overall article just not how it was written. I would have rewritten that quote to state that the profitable firms will be the ones that employ the best people regardless of race and sex. I agree that discrimination against black people and women is completely wrong and it deprives not only much needed employment to women, black people and those of the ethnic minority but it denies the companies of the potential of having a better quality workforce and fresh new ideas to better improve themselves financially.

[Winston Blake]

I did but explain to me this quote.

If so, discrimination is easy to detect in principle: Just note that the profitable firms will be the ones employing more women or workers from an ethnic minority

I agree the overall article just not how it was written. I would have rewritten that quote to state that the profitable firms will be the ones that employ the best people regardless of race and sex.

I agree with you. The fundamental assumption is that there are equal numbers and distributions of equally skilled people in all categories. For example, if you need 1000 people, and can pick between four pools of 1000 people each (White Male, Black Male, White Female and Black Female), then the company that picks 1000 white males is going to have a lower average than one which picks the best 250 from each pool.

However, if there's a pool of 5000 white males for this particular industry with a higher average skill level, then that argument no longer makes sense. Companies will only suffer significant costs in an ideal world, where the proportion of skilled women and population fraction of racial minorities become high enough.

[Turin]

I did but explain to me this quote.

If so, discrimination is easy to detect in principle: Just note that the profitable firms will be the ones employing more women or workers from an ethnic minority

Not to snipe from the sidelines, but come on. Well no shit, when you remove the sentence from all of its context, it sounds stupid. Read it with the rest of the paragraph it's in, and we see that it's simply a matter of Enigma failing at reading comprehension.

[Enigma]

I did but explain to me this quote.

If so, discrimination is easy to detect in principle: Just note that the profitable firms will be the ones employing more women or workers from an ethnic minority

Not to snipe from the sidelines, but come on. Well no shit, when you remove the sentence from all of its context, it sounds stupid. Read it with the rest of the paragraph it's in, and we see that it's simply a matter of Enigma failing at reading comprehension.

Dumbass, I understood the article but do not agree the wording of that quote. It just rubbed me the wrong way in an otherwise fine article.

[Turin]

Dumbass, I understood the article but do not agree the wording of that quote. It just rubbed me the wrong way in an otherwise fine article.

What "rubs me the wrong way" is that you apparently claim that you're able to comprehend the English language. Do you not see the If so at the start of that sentence? You know, the bit that ties it intimately with the entire paragraph above it? There's nothing to be "rubbed wrong" about with that sentence. You're just grasping at straws here.

[darthbob88]

Dumbass, I understood the article but do not agree the wording of that quote. It just rubbed me the wrong way in an otherwise fine article.

What "rubs me the wrong way" is that you apparently claim that you're able to comprehend the English language. Do you not see the If so at the start of that sentence? You know, the bit that ties it intimately with the entire paragraph above it? There's nothing to be "rubbed wrong" about with that sentence. You're just grasping at straws here.

I must agree with Enigma; the article does seem to suggest that, across the board, more women/minorities => More profits. It makes the assumption that women and minorities make up a significant enough portion of the labor pool that missing out on them will cause a noticeable loss of profits. If X% of the top candidates in a field are not white men, a company that employs Y% non-white-men will enjoy less profits if X > Y, because they will be missing out on a noticeable number of qualified candidates. However, so will a company which chooses to employ Z% non-white-and-male personnel, because it will miss out on the qualified white men.

In simplest terms, the article seems to assume that profits are directly and linearly proportional to minorities/women employed; Profits = K*X, where K is the percentage of non-white-men. From this model, profits would be highest when K is 100% minorities and women. It seems more realistic, from what I know, to assume profits are hyperbolically or parabolically proportional; profits are highest when the number of non-white-men employed is X%, and decrease as one moves away from X%.

[Turin]

darthbob88:
Aw, that's really cute the way you try to couch your argument in mathematical terms like a big boy. You want a cookie for that? Or better yet, how about you actually demonstrate that the article is saying what you think it is.

But discrimination should also show up in another way. Employers who prefer not to employ workers because of their sex or the color of their skin are likely to lose money: Employing stupid white men when you could be employing smart black women is not a profitable human-resources policy. Employers might nevertheless do this, either because they do not realize that their prejudices are costing them money, or because they do not care. If so, discrimination is easy to detect in principle: Just note that the profitable firms will be the ones employing more women or workers from an ethnic minority.

In other words, you lose out on qualified candidates because you're cutting out qualified candidates on the basis of race/sex. The only way this wouldn't be true is that there is a very large glut of qualified white male candidates far in excess for the demand for the position. Because in the real world qualified candidates are actually difficult to come by, this is a bullshit argument.

Additionally, and probably more importantly, you're assuming that "qualified" is a on/off condition, when it obviously isn't. A "super-qualified" candidate can have an even greater influence on profits than a minimally qualified one, but of course you have cut out a significant portion of your super-qualified candidates by rejecting them on the basis of race/sex. And because super-qualified candidates are even more rare than qualified candidates, you've multiplied the effect of the paragraph above.

[darthbob88]

Turin: Point. I was also assuming uniformity in qualified candidates, which would be unrealistic. Still, I do not like the implication that it is better to choose a woman or a minority over an equally qualified white man. It may be accurate, but it rubs me the wrong way. Point conceded.

[Turin]

Only you're still missing that there is no such implication made in the article. Where the hell are you and Enigma getting this except out of your festering rectums?

[Turin]

Apologies for the double-post, but just to come back and show that even your math was stupid, an example.

Number of positions and candidates

• Assume total # positions = 100
• Assume total # candidates = 1000
• Assume total # qualified candidates = 200 (20% of candidates)
• Assume total # super-qualified candidates = 50 (4% of candidates)

Candidate demographics

• Assume % of candidates who are ethnic minorities = 20
• Assume % of candidates who are women = 40
• Total % of candidates who are women/minorities = 52

Effects on candidate pool if you discriminate on race/gender

• # qualified candidates remaining if you eliminate women/minorities = 96
• # super-qualified candidates remaining if you eliminate women/minorities = 24

Hiring results (with discrimination)

• # qualified candidates hired if you eliminate women/minorities = 76
• # super-qualified candidates hired if you eliminate women/minorities = 24

Hiring results (without discrimination)

• # qualified candidates hired if you include women/minorities = 50
• # super-qualified candidates hired if you include women/minorities = 50

Which means that in this rather simplistic but fairly realistic example, you would double the number of super-qualified individuals in your organization by not engaging in discriminatory hiring practices.

The argument the article makes that because this is potentially such a large factor, one would in turn expect that the most profitable firms would be those who don't discriminate in the hiring of women and minorities (who, after all, together make up a majority of all possible candidates) and therefore have larger numbers of women and minorities in their firms.

[darthbob88]

Only you're still missing that there is no such implication made in the article. Where the hell are you and Enigma getting this except out of your festering rectums?

If so, discrimination is easy to detect in principle: Just note that the profitable firms will be the ones employing more women or workers from an ethnic minority.

More women/ethnic minorities => Fewer whites/men. Speaking as a white man, this does not please me.

More seriously, my argument is based on the assumption that "more minorities/women" equals, or can be reasonably taken to equal, discriminating in favor of minorities and women; this is obviously unreasonable, for much the same reason as discriminating in favor of white men.

As your example shows, color-blind hiring will produce more and better qualified candidates. However, if you favor women/minorities, you will get 26 super-qualified candidates
74 qualified candidates.
Granted, this is more and better than if you had favored whites and men, but it is still fewer than if you had chosen to hire based on qualification rather than discrimination. That was my argument; profits for a company which discriminates in favor of white men will be lower than one which does not discriminate, but so will a company which discriminates in favor of women/minorities.

[darthbob88]

Ghetto edit:
...As your example shows, color-blind hiring will produce more and better qualified candidates. However, if you favor women/minorities, you will get
26 super-qualified candidates
74 qualified candidates. ......

[Turin]

No shit, sherlock. And the article implies discriminating against white males... where exactly?

[darthbob88]

No shit, sherlock. And the article implies discriminating against white males... where exactly?

It does not directly suggest that, but I understood that phrase "more women and ethnic minorities" to imply "fewer of everyone else", IE white men. If discriminating for white men means discriminating against women and minorities, then it seems reasonable and logical that discriminating for women and minorities means discriminating against white men.

[Turin]

No shit, sherlock. And the article implies discriminating against white males... where exactly?

It does not directly suggest that, but I understood that phrase "more women and ethnic minorities" to imply "fewer of everyone else", IE white men. If discriminating for white men means discriminating against women and minorities, then it seems reasonable and logical that discriminating for women and minorities means discriminating against white men.

Non sequitor. Why don't you try to explain how "hiring women and minorities" equals "discriminating for women and minorities," numbnuts? Go ahead, this should be good.

[darthbob88]

No shit, sherlock. And the article implies discriminating against white males... where exactly?

It does not directly suggest that, but I understood that phrase "more women and ethnic minorities" to imply "fewer of everyone else", IE white men. If discriminating for white men means discriminating against women and minorities, then it seems reasonable and logical that discriminating for women and minorities means discriminating against white men.

Non sequitor. Why don't you try to explain how "hiring women and minorities" equals "discriminating for women and minorities," numbnuts? Go ahead, this should be good.

The article says that "more women and minorities would lead to more profits". Barring color-blind hiring, the only way to hire "more women and minorities" would be to discriminate in their favor.

Put it another way; had the article said that the more profitable companies would be those employing more white men, you would scream, and rightly so, that the article supported discrimination in favor of white men. Why should it be different when the article says women and minorities instead?

[darthbob88]

What I object to about that phrase, after thinking about it, is that it is misleading. It implies, or so I read it to imply, that a company which is 100% women and minorities would enjoy super profits, when this is clearly illogical. A more accurate phrasing would be "A company which employs a proportion of women and minorities representative of the candidate pool, evident of colorblind hiring policies, would enjoy greater profits than ones which hire fewer women and minorities." This should be more accurate, and less controversial.