December 12, 2002
Yes, Virginia, There Is Racial Discrimination in America

Alan Krueger writes about Marianne Bertrand and Sendhil Mullainathan's study of discrimination. The results are very depressing indeed--a lot of discrimination based on the racial association of the names on resumes, a lot more discrimination than I had thought such a study would find.

What's in a Name? Perhaps Plenty if You're a Job Seeker: ...To test whether employers discriminate against black job applicants, Marianne Bertrand of the University of Chicago and Sendhil Mullainathan of M.I.T. conducted an unusual experiment. They selected 1,300 help-wanted ads from newspapers in Boston and Chicago and submitted multiple résumés from phantom job seekers. The researchers randomly assigned the first names on the résumés, choosing from one set that is particularly common among blacks and from another that is common among whites. So Kristen and Tamika, and Brad and Tyrone, applied for jobs from the same pool of want ads and had equivalent résumés. Nine names were selected to represent each category: black women, white women, black men and white men. Last names common to the racial group were also assigned. Four résumés were typically submitted for each job opening, drawn from a reservoir of 160. Nearly 5,000 applications were submitted from mid-2001 to mid-2002. Professors Bertrand and Mullainathan kept track of which candidates were invited for job interviews.

No single employer was sent two identical résumés, and the names on the résumés were randomly assigned, so applicants with black- and white-sounding names applied for the same set of jobs with the same set of résumés. Apart from their names, applicants had the same experience, education and skills, so employers had no reason to distinguish among them.

The results are disturbing. Applicants with white-sounding names were 50 percent more likely to be called for interviews than were those with black-sounding names. Interviews were requested for 10.1 percent of applicants with white-sounding names and only 6.7 percent of those with black-sounding names...

## What's in a Name? Perhaps Plenty if You're a Job Seeker

By ALAN B. KRUEGER

HAT'S in a name? Evidently plenty if you are looking for a job.

To test whether employers discriminate against black job applicants, Marianne Bertrand of the University of Chicago and Sendhil Mullainathan of M.I.T. conducted an unusual experiment. They selected 1,300 help-wanted ads from newspapers in Boston and Chicago and submitted multiple résumés from phantom job seekers. The researchers randomly assigned the first names on the résumés, choosing from one set that is particularly common among blacks and from another that is common among whites.

So Kristen and Tamika, and Brad and Tyrone, applied for jobs from the same pool of want ads and had equivalent résumés. Nine names were selected to represent each category: black women, white women, black men and white men. Last names common to the racial group were also assigned. Four résumés were typically submitted for each job opening, drawn from a reservoir of 160. Nearly 5,000 applications were submitted from mid-2001 to mid-2002. Professors Bertrand and Mullainathan kept track of which candidates were invited for job interviews.

No single employer was sent two identical résumés, and the names on the résumés were randomly assigned, so applicants with black- and white-sounding names applied for the same set of jobs with the same set of résumés.

Apart from their names, applicants had the same experience, education and skills, so employers had no reason to distinguish among them.

The results are disturbing. Applicants with white-sounding names were 50 percent more likely to be called for interviews than were those with black-sounding names. Interviews were requested for 10.1 percent of applicants with white-sounding names and only 6.7 percent of those with black-sounding names.

Within racial groups, applications with men's or women's names were equally likely to result in calls for interviews, providing little evidence of discrimination based on sex in these entry-level jobs.

There were significant differences in interview-request rates among the nine names associated with black women, but not among the names within each of the other groups.

At the low end, the interview-request rate was 2.2 percent for Aisha, 3.8 percent for Keisha and 5.4 percent for Tamika, compared with 9.1 percent for Kenya and Latonya and 10.5 percent for Ebony.

Only part of this variability reflects chance differences resulting from sampling, although the authors have not been able to find a good explanation for the wide range. Thus it is important that the names chosen for black women were not uncommon; they represent 7.1 percent of all names listed on Massachusetts birth certificates for black girls from 1974 to 1979.

The 50 percent advantage in interview requests for white-sounding names held in both Boston and Chicago, and for both men and women.

This discrepancy complements findings from earlier studies in which researchers sent a small number of matched black and white "auditors" to apply for jobs in person. Typically, though not always, the black job seekers were less likely to be invited for an interview or offered a job.

Those findings, however, were criticized because the applicants knew the intention of the study and might have behaved differently. In addition, the auditors might not have been well matched with the jobs in question; they could have been overqualified or underqualified.

Professors Bertrand and Mullainathan's study is less susceptible to these concerns. First, they used a large number of names and inanimate résumés. Second, the job openings involved administrative, sales, clerical and managerial positions, and they submitted résumés patterned after real résumés of people who were actually seeking similar jobs.

Their most alarming finding is that the likelihood of being called for an interview rises sharply with an applicant's credentials — like experience and honors — for those with white-sounding names, but much less for those with black-sounding names. A grave concern is that this phenomenon may be damping the incentives for blacks to acquire job skills, producing a self-fulfilling prophecy that perpetuates prejudice and misallocates resources.

Two main theories explain labor market discrimination. One, known as taste-based discrimination, posits that employers — or customers, co-workers or supervisors — have a preference against hiring minority applicants, even if they know they are equally productive.

The other, known as statistical discrimination, assumes that employers personally harbor no racial animus but cannot perfectly predict workers' productivity. In this case, an employer assessing an applicant would assign some weight to the average performance of the person's racial group, instead of basing the judgment solely on the individual's merits.

A difference between these models is that employers sacrifice profits to indulge in taste-based discrimination, while, in principle, statistical discrimination, if based on accurate information, can help the bottom line. Professors Bertrand and Mullainathan cannot distinguish between the models — and both may be applicable — but they suspect that their finding that employers in heavily black areas of Chicago are less likely to discriminate against black-sounding names augurs for taste-based discrimination.

Nevertheless, either rationale for discrimination is illegal and prohibited.

"That which we call a rose," Juliet said, "by any other name would smell as sweet." An organization like the Civil Rights Commission or the Equal Employment Opportunity Commission could perform a service if it routinely monitored discrimination by conducting audit studies similar to Professors Bertrand and Mullainathan's.

Posted by DeLong at December 12, 2002 12:17 PM | Trackback

Email this entry
Email a link to this entry to:

Message (optional):

And I want to thank my father right now for giving me the particularly white name of Wesley.

Thank you, thank you, thank you.

Posted by: on December 12, 2002 01:41 PM

but isn't race a proxy for class? i can't imagine the results being all that different that one of the groups been given traditional poor white trash names like earl and billy bob.

Posted by: kit on December 12, 2002 01:44 PM

but isn't race a proxy for class? i can't imagine the results being all that different had one of the groups been given traditional poor white trash names like earl and billy bob.

Posted by: kit on December 12, 2002 01:45 PM

but isn't race a proxy for class? i can't imagine the results being all that different had one of the groups been given traditional poor white trash names like earl and billy bob.

Posted by: kit on December 12, 2002 01:45 PM

Even from the complete NY Times article, it's hard to tell whether the study was legitimately designed or not.

It appears to me that the authors may have projected some of their perspectives and opinions into the study - whether intentional or unintentional - that may have strongly biased the results.

As alluded to by kit, the names selected for black women, though common recently, were not traditional names. Is it possible that employers have a bias towards traditional solid sounding names?

Think about branding. Lots of time and effort and attention goes into selecting a brand name that brings forth some idea of what the product does well. It's not racial, it's just branding.

For example, a neighbor of mine named his daughter "Fleetwood". Were I to get a resume from a Fleetwood and a Karen and I only had time for one interview, I'd go with the Karen. Every time. BTW, Fleetwood is white.

So anyway, unless the non-black names included a reasonable sample of non-traditional names, I don't think it's fair to conclude race was the factor driving results in this study.

Posted by: Anarchus on December 12, 2002 01:59 PM

Professor Krueger is a most careful analyst. The study appears to show again how easily we stereotype and hard slowly racial stereotypes are changed.

Posted by: on December 12, 2002 02:08 PM

"Is it possible that employers have a bias towards traditional solid sounding names?"

"Were I to get a resume from a Fleetwood and a Karen and I only had time for one interview, I'd go with the Karen. Every time. BTW, Fleetwood is white."

I believe that this is of far greater importance. My hunch is that the results would have been similar if "white trash" type names were used. There's little doubt that a conventional sounding name will open more doors. Is this fair? Nope, but let's not jump to an invalid conclusion that racism is involved.

Posted by: David Thomson on December 12, 2002 02:19 PM

I recall a dustup in the midwest a couple of years ago, when a school district proposed renaming a local high school after Martin Luther King. The white parents flipped out, fearing college admissions offices would stereotype its students and/or assume the grades were less-than-meaningful.

Also, I didn't see the cited study mention address as an issue. It seems reasonable that a LaTisha Williams from Greenwich, Connecticut, might be perceived differently than a LaTisha Williams from the Bronx.

Posted by: George Zachar on December 12, 2002 02:58 PM

"I recall a dustup in the midwest a couple of years ago, when a school district proposed renaming a local high school after Martin Luther King. The white parents flipped out, fearing college admissions offices would stereotype its students and/or assume the grades were less-than-meaningful."

The sad thing is that the well meaning Liberals who dropped the standards in the universities for the purposes of affirmative action have inadvertently caused enormous harm. Unfortunately, it is fair to wonder if a black person's degree is worth the paper it is printed on. I have long advocated affirmative policies premised upon picking a minority in certain circumstances---if all the other factors were equal! Need I point out that this is not what usually occurred.

Posted by: David Thomson on December 12, 2002 03:13 PM

Anarchus's point is a good one, and I'd frankly be amazed if there aren't already studies on the books documenting that exact finding (traditional vs. non-traditional names).

Given that, I'm guessing that this factor is discussed in the actual source journal article, and that it was partialed out in at least some of their analyses and still yielded an effect.

Anyone have a cite for the original source?

Posted by: rufus on December 12, 2002 03:30 PM

No David, it is not fair to wonder if a black person's degree is worth the paper it's printed on. That's YOUR bias and racism speaking. You're assuming that even if a person's acceptance to a school was based on something besides pure academic performance (like race, or musical ability, or parental donations or athletics, or being the child of an alum), the fact that they graduated (with a particular grade point average) isn't worth anything. You're assuming that they spent 4 years in college and didn't earn a single grade, didn't learn, didn't deserve to graduate, all because of the color of their skin.

Posted by: lisse on December 12, 2002 03:41 PM

The sad thing is that the well meaning Liberals who dropped the standards in the universities for the purposes of affirmative action have inadvertently caused enormous harm. Unfortunately, it is fair to wonder if a black person's degree is worth the paper it is printed on.

This is absolutely despicable.

Posted by: Dennis O'Dea on December 12, 2002 03:42 PM

I remember once telling a VP where I worked that I had a resume from someone who would have been perfect as a test engineer for his department (the guy's experience wasn't a fit for mine), and he said, "Great! What's his name?" and I said, "Mohammed (something)" and he said "not interested." I'm pretty sure, having gotten to know the VP, that he would have done the same thing for Tyrone or Tawana.

The same company had a sales VP who spent most of his spare time on tirades about "lazy f*cking n*gs." Mr. High-Powered, had a big house in Atherton. He was a pig, and there are many people like him out there in this country, whether anyone wants to admit it or not.

Affirmative action at universities was almost always about admissions, not ongoing grading standards. There's a big difference between being admitted to a top school and graduating from one.

Posted by: Peter MacLeod on December 12, 2002 03:51 PM

Did David Thomson really just write this or am I dreaming;

"Unfortunately, it is fair to wonder if a black person's degree is worth the paper it is printed on"

Posted by: James on December 12, 2002 03:53 PM

>> Brad and Tyrone, applied for jobs from the same pool of want ads and had equivalent résumés <<

The above makes me suspicious. Just what is an equivalent resume? To make the obvious point, black college graduates are much more likely to have been admitted under lower academic standards/test scores, and employers know this.

Further, some groups of people are dangerous to a deep pockets employer, as they are more likely to file a discrimination lawsuit if their employment doesn't go well. Again, this is well known to employers.

But let's look at the real world for a change (and outside the atypical cities of Boston and Chicago). Productive employees are hard to find, and resumes aren't much use in identifying them. So I never have had much use for them except as a handy way to have a phone number on file.

My very small labor force has been recruited personally by observing people working at jobs that I (usually correctly) think don't pay much. If I spot someone who displays skills I think could be useful, I make a point to get to know him/her. I've hired a lot of people and raised their incomes this way. Most jobs don't require skills that can't be taught on the job, if the employee has a good attitude.

We greedy entrepreneurs have our uses, guys.

Posted by: Patrick R. Sullivan on December 12, 2002 03:58 PM

That's right patrick. White sounding names have a 50% better chance of being called for an interview than black sounding names is really just indicative of entreprneurs using their talents, being innovative, and skillfull, and helping the economy.

Posted by: Dennis O'Dea on December 12, 2002 04:01 PM

FOUND IT !!

Go here:

http://www.econ.yale.edu/seminars/apmicro/

and scroll down to Ms. Bertrand's presentation on December 4th - there's a pdf file link there.

Posted by: Anarchus on December 12, 2002 04:08 PM

A preference for traditional sounding names?

Maybe, but it's a stretch.

First, from the NYT article it is possible to infer that the study authors paid some attention to the popularity of the names they used, and to the variation within groups.

Second, why is Kristen a "traditional" name, while Tyrone isn't? Is it possible that our notion of what is and isn't traditional has a racial tinge? Would you consider "Rufus" traditional? William the Conqueror had a son by that name. Is "Willie" less traditional than "Brad?"

It's easy to think up possible alternative explanations, but to claim there's not some racial discrimination going on here is far-fetched.

Posted by: Bernard on December 12, 2002 04:10 PM

The study involved sending resumes for entry level administrative and clerical jobs. Not brain surgery, but jobs that would require a resume. Presumably not fast food, not mall store sales persons.

I cannot even comment on Patrick's assertion that he goes around evaluating the productivity of random white people to raise their salaries. Whatever.

Posted by: lisse on December 12, 2002 04:16 PM

Went through the study quickly.

The author(s) dance around the issue of unusual versus traditional names, but hide behind the fact that all the names came from actual SSN data in ratio's of popularity. I would note that since one factor they used was the ratio of white name frequency to black name frequency, they inadvertently biased the study away from less modern black names - because the most modern black names would have the least overlap with white names.

I'm not an expert on naming, but I believe there's been a strong trend in the African American community over the past 20-30 years to select names for their children that were extremely rare in the past. Aisha and Keisha are two names mentioned in the study which have become popular recently. I could be wrong, but I don't think those names have been used much historically prior to 30 years ago.

Rather than relying entirely on SSN information and ratios for name selection, it would have been interesting if the researchers had thrown in a few funky white names (Sierra and Chastity come immediately to mind) and a few traditional black names (names that were in common use in the black community prior to 30 years ago) and analyzed that data.

Posted by: Anarchus on December 12, 2002 04:24 PM

"Yes, Virginia, There Is Racial Discrimination in America."

Who would have thought it?

I don't know if anyone saw the John Stossel (ABC) piece on discrimination on the basis of looks, but:

1) They sent in both men and women for job interviews, where both had essentially identical resumes. But in each case, one was really good looking, and one was only average looking. The really good looking person got hired every time (as I recall...it's been over 6 months since I saw it).

This was true, even if the interviewer was the same sex as the interviewee. I remember one case where the (heavyset, balding) stockbroker was interviewing a Steve Garvey-type (that will date me) good looking guy. The heavyset, balding stockbroker said, when he was interviewed, "He (the Steve Garvey guy) just LOOKED like a stockbroker."

2) They had two woman teachers come into a kindergarten or 1st or 2nd grade class. One was drop-dead gorgeous. The other just OK.

What did the kids say about the drop-dead gorgeous one? "She was nicer." "She was smarter." Why was she "smarter"? "Wellllll...because she was prettier."

3) They had 3 guys, one 5'7", one 5'4", and one 5'0". They put one or the other in a lineup with taller guys, and asked women to rank the lineup by who they most wanted to date. No matter HOW great they made the short guys sound..."He's a multi-millionaire owner of his own company, flies his own Lear jet, loves to fly to Europe to go dancing..."...the short guy was always at the bottom of the list in the women's choices.

As one woman said, when interviewed about why she didn't choose the short guy, despite his incredible "credentials," she said, "Well, maybe if all the other guys had been serial killers..."

Posted by: Mark Bahner on December 12, 2002 04:33 PM

"Did David Thomson really just write this or am I dreaming;

"Unfortunately, it is fair to wonder if a black person's degree is worth the paper it is printed on"

Nope, you are wide awake. Please don't shoot the messenger. My only responsibility is to the truth. Why blame me? Take your complaints to the Liberals who caused this mess. The following item might be found of interest:

"...Through affirmative action, a number of black students were accepted to Harvard with SAT scores and academic performance below that of the average Harvard student, the Globe reported."

http://www.browndailyherald.com/stories.cfm?S=2&ID=5262

Posted by: David Thomson on December 12, 2002 04:45 PM

"Harvard's pluses and minuses are likewise discussed candidly. While Harvard "still retains many of the best-known scholars and thinkers in the country," there are also "real problems at Harvard" -- "the core curriculum is large and undefined; grade inflation has become outrageous; and political correctness is to many there a higher calling than intellectual integrity."

http://www.jewishworldreview.com/cols/sowell100298.asp

Posted by: David Thomson on December 12, 2002 04:54 PM

Brilliant. But isn't the practice of classifying names as "black" and "white" a bit too politically incorrect to actually put into practice by the Equal Employment Opportunity Commission. Are the results of the study reproducible? The study's results show entry level jobs. What about higher up jobs like, say, pre-business school jobs? The myth in the Ivy League is that being black from a place like Harvard, Yale, Princeton, etc. provides a boost for those more lucrative jobs. . . .

Posted by: Bobby on December 12, 2002 04:54 PM

lisse and Peter MacLeod's point that graduating is (less susceptible to affirmative action, and thus) more meaningful than admittance is quite strong. One needs to add, however, that degrees from the same school (with the same GPA) from different departments can be wildly different in their implications about their degree-holder.

Regarding lisse's later post, I think Patrick's point was that he was not looking specifically for white people, but people (of any race) he could pay slightly more than they earn now, but who would be especially productive.

Patrick, how certain can you be that you do not have some unconscious bias in your process that makes you less likely to select equally high productivity/wage black candidates than white candidates? There are several steps in the process you describe that are potentially subject to unintentional bias.

Posted by: Tom on December 12, 2002 04:58 PM

David Thompson, don't you have a home to go to?

Posted by: on December 12, 2002 05:29 PM

"But isn't the practice of classifying names as "black" and "white" a bit too politically incorrect to actually put into practice by the Equal Employment Opportunity Commission."

It's also impractical. Rufus Sewell is a well known white British actor, but many might incorrectly think that he is black. The only way that the EOC could enforce such a stupid and reprehensible law is if we restricted certain names to particular racial groups. Oh darn it, that won't work either. What do you do with a Tiger Woods? Why does he have to make things so complicated?

Posted by: David Thomson on December 12, 2002 05:36 PM

"...Through affirmative action, a number of black students were accepted to Harvard with SAT scores and academic performance below that of the average Harvard student, the Globe reported."

Ok, I would expect half to be at or below average.
What's the point?

Posted by: Brian on December 12, 2002 05:47 PM

"...Through affirmative action, a number of black students were accepted to Harvard with SAT scores and academic performance below that of the average Harvard student, the Globe reported."

Wanna bet that through legacy admissions and athletic preferences, a number of white students were accepted to Harvard with SAT scores and academic performance below that of the average Harvard student? Since it is an "average," one would expect many students to be below the average.

"Harvard's pluses and minuses are likewise discussed candidly. While Harvard 'still retains many of the best-known scholars and thinkers in the country,' there are also 'real problems at Harvard' -- 'the core curriculum is large and undefined; grade inflation has become outrageous; and political correctness is to many there a higher calling than intellectual integrity.'"

And your point is? Is this another attempt to hijack the thread into a discusison of the evils of liberalism?

Posted by: nameless on December 12, 2002 05:50 PM

Another hypothetical take on the address issue:

Three women, LaTisha Williams, Sophie Schwartz and Jane Smith, all sharing the same Eastern Parkway, Brooklyn, NY address, apply for a job as a bookkeeper.

That address is in a famously polyglot neighborhood, Crown Heights, populated by Afrrican-Americans, Orthodox Jews, and gentrifying yuppies. Who gets the call?

A person who is more likely than the other two to sue for racial discrimination, the one who is more likely to require many extra days off for Jewish holidays, or someone who seems to bear neither of those risks?

It's a rational question a personnel manager would ask.

Posted by: George Zachar on December 12, 2002 06:19 PM

“And your point is? Is this another attempt to hijack the thread into a discussion of the evils of liberalism?”

I guess I’m just a mean spirited thread hijacker who finds it peculiar that you even think we can discuss this issue without dealing with the unintended consequences of well meaning Liberal policies. And yes, the good old white guys hypocritically made sure that their stupid kids were well taken care of by the system. Nevertheless, the dropping of standards for minority students inevitably cheapened their degree. Why should that be so hard for you to comprehend? Please also note that I publicly champion well thought out affirmative action policies.

Posted by: David Thomson on December 12, 2002 06:24 PM

"I guess I’m just a mean spirited thread hijacker who finds it peculiar that you even think we can discuss this issue without dealing with the unintended consequences of well meaning Liberal policies."

I understand the point your are trying to make regarding affirmative action. I don't understand what other "liberal policies" you are referring to or why you think the alleged problems at Harvard are relevant to this discussion.

Posted by: nameless on December 12, 2002 06:42 PM

You can get the data on the distribution of names from the census site. What I found facinating is that they follow a power-law distribution. Those are generally a sign of perferential attachment. It's interesting to see experimental data that suggests why people might tend to link up with an already popular name.

Here is a chart of the distribution for last names.
http://enthusiasm.cozy.org/archives/2002_06.html#000012

The full paper:

http://www.econ.yale.edu/seminars/apmicro/am02/bertrand-021204.pdf

is just excellent. Your better off with a better zipcode too. It reports that my delightful and brilliant friend Kareem has only a 4.7% chance of a call back when he sends in a resume. Hakim, who we have intrusted with a very important peice of work, has only a 7.3%. Brad will be happy to know he can look forward to 15.9% of the resumes he sends in will get him a callback. I may need to rename my children.

Just last week I was told "Ah 'Ben' that's a money name." Hasn't worked so far.

Posted by: Ben Hyde on December 12, 2002 06:49 PM

One of the best predictors of Presidential success is . . . . . . height. The taller U.S. Presidential candidate wins something like 90% of the time.

People are funny creatures.

Posted by: Anarchus on December 12, 2002 07:03 PM

Anyone who thinks that racial discrimination doesn't exist should go into a fast food restaurant. Why is the only white males working there the manager? Why does that seem "normal."
As a reasonable attractive white male, I'm sure some of my success at my job is because people perceive that I "look" the part. And there is no doubt that race and sex plays a role in looking the part. It just is the way it is. We've done some things to address it, but its a long, long way from fixed.

Posted by: pj on December 12, 2002 07:05 PM

Why is it so hard to believe that common every-day racism is alive and well? Blacks are lazy; blacks are more likely to sue. It's fair to disbelieve the academic credentials of a black-named resume. My god. The posts right here establish the point that the study was trying to make.

First we had Jim Crow laws; then in response to busing we get white flight. Now, racial tipping is a huge issue in many schools drawing from mixed communities. The drumbeat echos on from the Civil War -- blacks are unequal. Once upon a time, the explanation was genetics. Now, apparently, it is the fault of (beware, dirty word to follow) liberals.

Are Anarchus and David Thomson and Patrick Sullivan and George Zachar racists? It depends on how you define the term. I don't see any of them advocating a return to Jim Crow. But they all show that very modern subtle racism that says that blacks are still unequal, but this time its not their fault.

Bah. It stinks. When blacks kids are even vaguely close to white kids of similar economic background in equality of opportunity, then you can tell those damn liberals to back off. Until then, the ghost of slavery and Jim Crow hovers over people like David Thomson. He is justifying disparate treatment based purely on race and that is a perfectly workable definition of racism.

I knew from earlier posts, David, that you hated the French. Its refreshing to see your true colors about your feelings toward minorities.

Posted by: FDL on December 12, 2002 07:15 PM

'To make the obvious point, black college graduates are much more likely to have been admitted under lower academic standards/test scores, and employers know this.'

What they have to do to graduate is the same as white students. This is also known to employers.

Unless you really think a significant component of a degress is a SAT score proxy, that is.

'One needs to add, however, that degrees from the same school (with the same GPA) from different departments can be wildly different in their implications about their degree-holder.'

Not in this study; they had the same resume.

'A person who is more likely than the other two to sue for racial discrimination, the one who is more likely to require many extra days off for Jewish holidays, or someone who seems to bear neither of those risks?

It's a rational question a personnel manager would ask.'

It's also illegal.

Posted by: Jason McCullough on December 12, 2002 07:34 PM

“I knew from earlier posts, David, that you hated the French.”

My harsh denunciation of the French is most assuredly not racially based. My own ancestors immigrated from Germany--and there is always the possibility that they lived in an area that is now part of France. No, my problems with the French are due almost solely to the failure of their Revolution. The American revolution was a success because it was premised upon a realistic understanding of human nature--and stressed the mandatory need for check and balances. The French revolutionaries regrettably advocated pure democratic principles. This has resulted in much horror and suffering. Oh by the way, I also have a lot of contempt for German philosophy. Martin Heidegger, Friedrich Nietzsche, and Carl Schmitt are among the most evil thinkers who ever lived.

Today the French parasite off the United States and will not hesitate to stab us in the back. This is just a cold fact of life.

Posted by: David Thomson on December 12, 2002 07:43 PM

"Bah. It stinks. When blacks kids are even vaguely close to white kids of similar economic background in equality of opportunity, then you can tell those damn liberals to back off. Until then, the ghost of slavery and Jim Crow hovers over people like David Thomson. He is justifying disparate treatment based purely on race and that is a perfectly workable definition of racism."

You are obviously not paying attention. I advocate affirmative action policies! How much clearer do I have to be regarding this point? Liberals, though, have made a serious mistake in lowering the standards for minority students. I’m all for selecting the minority candidate, in many circumstances, when everything else is equal. Unfortunately, this has not always been the case.

Posted by: David Thomson on December 12, 2002 08:00 PM

“I knew from earlier posts, David, that you hated the French.”

I almost forgot to add something of significance. It is my adamant conviction that the United States may very well have better relationships in the future with a number of nonwhite nations instead of the members of the so called European Union. India, Turkey, and even Iran may very well soon be closer to us. Those white guys in Europe are going downhill fast because of their silly socialism. They be too far gone to be saved. The same also holds true for the Canadians.

PS: We are also already closer to Israel than with the European union folks.

Posted by: David Thomson on December 12, 2002 08:39 PM

Studies like these shows that discrimination exists, but how much does it actually contribute to lesser earnings for blacks? Other research has shown that most, or in some cases, all of the earnings gap between whites and blacks or Hispanics can be explained by differences in skills.

Posted by: tc on December 12, 2002 09:02 PM

Wait a minute. Ebony got above average. Few names could be as obvious as Ebony. How do you think Ardvard Ramadan would do. I am extremely liberal but Aisha what kind of name is that. Sounds Indian to me. I think this whole study is bogus. And I do believe in discrimination. People are more afraid of what they do not understand than anything else.

Posted by: Bruce Ferguson on December 12, 2002 09:36 PM

Does anyone agree that if you get a number of resumes that show the exact same qualifications that you should see them all? Or, if that is impractical, that you should make a blind decision about which ones to see? Is it not reasonable that any decent human being would not choose to interview a worker on the absurd, nonsensical basis of their name? What kind of people make decisions based on something that stupid?

What an incredibly depressing thread. Jews take too many days off, black people's college diplomas aren't worth the paper they're printed on, the way your name "sounds" is a good reason not to call you for an interview, certain people are presumed more likely to sue you for discrimination so you are wise to pre-emptively discriminate against them, therefore proving that they ARE subject to discrimination. wow.

Liberals (like Richard Nixon) "lowered standards" with affirmative action, but letting George W. Bush have a slot at Yale that could have otherwise gone to a smart kid doesn't.

And none of this is based upon irrational prejudice. In fact, those who defend such "reasonable" thought processes are the very people who believe so strongly in individualism that any kind of group designation is anathema. Unless, of course, one is reasonably deciding before even meeting a person if they will want extra days off or will be likely to sue them. This is the "real world" where liberals have caused all these problems by foolishly attempting to mitigate some of the unexamined, reflexive bigotry and stereotyping... that we see on this thread.

Posted by: digby on December 13, 2002 12:08 AM

If you have the military test AFQT as a proxy for skills the wage ratio of a black to white person with same score is about .95. And I think that an Oaxaca decomposition of a Mincer earnings function implies the ratio is somewhere from .87 to .9 depending on which variables you control for. Of course both these measures of discrimination have problems, but the study cited by Krueger seems to imply that the wage differentials I just cited are off by an order of magnitude.

Posted by: Bobby on December 13, 2002 12:12 AM

The authors don't gloss over the name question though, they control for it with a separate manipulation. The key finding is that WITHIN racial name groups, "white" names benefit from having more experience while "black" names do not.
This interaction of the two effects demonstrates that the name factor alone is not driving the results.

In related news, someone voiced skepticism that the "name issue" could be a driving factor here.
I would guess that this person is lucky enough not to be involved in behavioral research :)
Little variables like these are the bane of your existence, and you spend a lot of your time trying to control for them (and no matter how many you control, some reviewer finds another one).

As I said though, given that fact, I was going to be very surprised if the authors hadn't controlled for it in some way, and they in fact did.

Posted by: rufus on December 13, 2002 12:23 AM

Discrimination exists. That doesn't mean that all studies that claim to show discrimination are valid.

Posted by: Ehud on December 13, 2002 04:25 AM

"Liberals (like Richard Nixon) "lowered standards" with affirmative action, but letting George W. Bush have a slot at Yale that could have otherwise gone to a smart kid doesn't. "

How many times do I have to repeat myself? I have already addressed that point. The guilt over the past hypocrisy likely opened the door to a lot of this nonsense.

Posted by: David Thomson on December 13, 2002 04:47 AM

“What an incredibly depressing thread. Jews take too many days off”

I believe it was a Jewish poster who pointed out this practical consideration that an employer might dwell upon. There are often awkward situations that have to dealt with in the real world. I don’t want to put words into the mouth of another poster, but this individual obviously wasn't talking about Jews per se. He was likely referring to those of a more orthodox persuasion who would ask not to work on Friday evenings and all day Saturday. Does anyone have a perfect answer to such a dilemma?

“black people's college diplomas aren't worth the paper they're printed”

Blame that on the Liberals. They are the ones who are mostly responsible for dropping the standards.

“the way your name "sounds" is a good reason not to call you for an interview, certain people are presumed more likely to sue you for discrimination so you are wise to pre-emptively discriminate against them, therefore proving that they ARE subject to discrimination. wow.”

Yes, this is what real people do in the real world. They usually do not have a tenured position in the make believe atmosphere of a university. Any employer who does not worry about the future possibility of an employee law suit---is a damn fool! Your self righteousness is part of the problem, and not part of the solution. People like you have added greatly to our national troubles. You refuse to listen to their very valid concerns and instead prefer to simply accuse them of "unexamined, reflexive bigotry and stereotyping."

Posted by: David Thomson on December 13, 2002 05:14 AM

EXCUSE ME.

FDL, please defend this. "Are Anarchus and David Thomson and Patrick Sullivan and George Zachar racists? It depends on how you define the term."

Under what possible definition do you consider me a racist? AND THIS JUST RELATES TO ME - I don't know who those other people are, what they've done and I don't care - but I do care about my reputation, which you've attempted to smear.

THERE IS NOTHING I wrote above which doesn't relate specifically to this study - which is very flawed - how on earth does criticizing a pro-race-bias study make me a racist?

I could easily argue that it's racist to put these issues off limits to fair, constructive criticism - that racial issues and African Americans are so delicate that they can't stand up to normal scrutiny. Which is utter nonsense.

PS: I have an adopted mixed race child and I've written four letters to my senators demanding Trent Lott's resignation. FDL: STUFF IT.

Posted by: on December 13, 2002 05:52 AM

To FDL: Can you provide data showing blacks and whites are equally likely to file racial discrimination lawsuits?

Old joke: How do you know when a liberal is backed into a corner? He calls you a racist.

Posted by: George Zachar on December 13, 2002 06:03 AM

I think it is time for an irrelevant post. If your choice of name determines how far you advance in America, the smart thing to do is go by first initial, middle name.

H. Ross Perot -- Before losing two presidential elections, he was a very successful businessman.

M. Steve Forbes -- despite his creepy, creepy looks he has run a successful magazine, and managed not to fritter away his family's fortune -- despite running for president twice.

J. Danforth Quayle -- before becoming a comedic goldmine, he was a rising star in the Republican party. SO much so, that an elder party statesman George Bush chose him as a running mate.

J. Bradford DeLong -- needs no introduction.

Posted by: Sterling on December 13, 2002 06:04 AM

The comment defending Anarchus above was by me (Anarchus). Don't know why the name didn't print out the first time.

Posted by: Anarchus on December 13, 2002 06:05 AM

“What an incredibly depressing thread. Jews take too many days off”

I believe it was a Jewish poster who pointed out this practical consideration that an employer might dwell upon. There are often awkward situations that have to dealt with in the real world. I don’t want to put words into the mouth of another poster, but this individual obviously wasn't talking about Jews per se. He was likely referring to those of a more orthodox persuasion who would ask not to work on Friday evenings and all day Saturday. Does anyone have a perfect answer to such a dilemma?

Yes, I am Jewish, and yes, I was talking about Orthodox Jews who, in addition to having to leave work around 3 every Friday, also take off many days for legitimate holidays that do not coincide with the secular business calendar.

I know folks who have had to factor this into their personnel decision making.It is childish to pretend such issues do not influence hiring. And pointing out a real world fact does not make one a racist or an anti-semite.

Posted by: George Zachar on December 13, 2002 06:13 AM

I don't understand why the perception that certain classes of applicants are more likely to litigate in case of dismissal isn't an important factor in this result about applications. This isn't just racism; it's also rational action. I've also seen a graph showing workforce participation by the disabled since the passage of the big Bush disability protection law. The graph showed a gradual decrease in participation. I think legal remedies may actually be counterproductive.

Posted by: JT on December 13, 2002 06:45 AM

>> I cannot even comment on Patrick's assertion that he goes around evaluating the productivity of random white people to raise their salaries. Whatever.<<

Here is what I actually wrote, and note the complete absence of a mention of white people:

>> [I] recruited personally by observing people working at jobs that I (usually correctly) think don't pay much. If I spot someone who displays skills I think could be useful, I make a point to get to know him/her. I've hired a lot of people and raised their incomes this way.<<

"lisse", you appear to be a racist! Not to mention that you seem to be unaware of the lesson Adam Smith taught us over 2-1/4 centuries ago about not expecting to get our meal from the baker and the butcher's altruism.

Posted by: Patrick R. Sullivan on December 13, 2002 07:01 AM

I shouldn't even enter into this thread, in which many of the usual suspects have amply proven the study to be well-grounded in reality.
First, a quibble: what percentage of American Jews are observant Orthodox? Surely the odds of Sophie Schwartz (who may be Greek married to a Jew for all our HRC manager knows) taking a lot of days off are not much different from Jane Smith being a sickly. Besides, who wouldn't want someone who'll gladly come in on Christmas, not leaving early on Xmas Eve or Good Friday? Then there's the possibility of an employee who regularly brings in chicken soup! OK, ok, I'm sorry. This thread is just too goofy.
Tyrone from Harvard isn't good enough for David. Would he hire Sterling Carlisle from Harvard? That guy's obviously a legacy, and _his_ degree isn't worth the paper it's printed on. What about Gene Hoffman? Might be Abby's kid! An anarchist in our midst!
Who knew that Trent Lott's fan club would find a home here?

Posted by: JRoth on December 13, 2002 07:11 AM

>> That's right patrick. White sounding names have a 50% better chance of being called for an interview than black sounding names is really just indicative of entreprneurs using their talents, being innovative, and skillfull, and helping the economy.<<

reminds me of one of the most entertaining half-hours of television I've ever seen; Thomas Sowell's appearance on Meet the Press approximately twenty years ago. The panelists actually ran out of questions to ask him, because they couldn't comprehend the rather simple message he had to convey.

I believe it was Marvin Kalb who finally asked: "So, you think racism is just going to go away?". To which Sowell responded along the lines of: "On the contrary, I think there will be racism as long as there is oxygen on the planet. What I AM SAYING, is that I originally thought I could quantify the effect racism had on such things as incomes of minorities, but I couldn't find any such effect.".

It apparently being such a strong stereotype in the minds of Kalb and company that there WAS such effect, they were unable to process what Sowell was telling them in plain English. We seem to have a similarly strong stereotype from some in this thread.

Posted by: Patrick R. Sullivan on December 13, 2002 07:18 AM

I'm still mystified. Our very own Trent Lott, David Thomson, says this;

"Unfortunately, it is fair to wonder if a black person's degree is worth the paper it is printed on."

Then tries to justify it by saying ENTRY standards are lower (whilst not offering much proo, admittedly). What's that got to do with degree results (i.e EXIT standards).

I'm sure it's something to do with the Europeans parasiting off the US, but I can't quite work it out how.

Posted by: Matthew on December 13, 2002 07:21 AM

Morning, all. I think, having run through this thread, that I have an as-yet unaddressed question about this study. Now, I have only read the summary NY Times article, so whoever read the actual study may be able to help me here.

The authors seem to be assuming that the correct callback ratio for interviews should be 50/50. Fewer blacks being called back is evidence of racism.

Well, it may be evidence of non-colorblindness in the hiring process. However, my (laymens) understanding of EEOC guidelines is that a company's workforce should reflect the ethnic composition of the region. So, suppose Massachusetts has two whites for each black. A company might be calling back and hiring two whites for each black - no racism at all, EEOC approved! And if all companies did this, we would have an environment with no overt hiring racism, but hideous name-oriented racism based on studies like this.

As I said, I don't know how the authors controlled for this.

Now, may I go into the bonus round? The NY Times had an article earlier this week where some social psychologists had an interesting idea for testing bias. Test subjects were shown photos of white and black men holding either guns or innocent objects like soda cans or cameras. If the photo showed a man with a gun, the test subject pushed a button to "shoot" him.

Hmm. Whites were quicker to shoot blacks than whites, and shot more unarmed black men than whites. This was true regardless of how people did on questionaires designed to test their tolerance.

As a bonus, the researchers mentioned preliminary results where black people are doing the shooting. Hmm, again. Now it is blacks who are quicker, rightly or wrongly, to "gun down" blacks.

Let me repeat, for the unusually sensitive in the readership, no black or white people were actually shot during these tests. I apologize if I seem unduly callous in reporting these results.

I am having a computer glitch, but I will try to post a link to the Times article shortly.

Posted by: Tom Maguire on December 13, 2002 07:23 AM

OK, Times article on racism as tested in video games:

http://www.nytimes.com/2002/12/10/health/psychology/10RACE.html?pagewanted=print&position=top

Posted by: Tom Maguire on December 13, 2002 07:25 AM

From a statistical standpoint, here's the issue.

At the very start of the study, the authors assumed that by selecting the most frequent white names and the most frequent black names, and then adjusting to reduce overlap, that they had created a theoretical "pool" of candidates that differed only in the "whiteness" or "blackness" of their names.

I believe that their assumption is false, and that there's at least one other meaningful statistical difference (call it modernity of name, for lack of a better term) that influences their results. That's all I've ever written.

It would have been interesting to see whether equally modern (or redneck) white-identified names were discriminated against, but the authors didn't look at this factor.

As for whether or not it's good for society that employers prefer people named David to Tyrone, of course it's not, but people are very irrational about many similar things. As mentioned before, the height of political candidates is a very strong predictor of election success? That's not right, either, but it doesn't mean that people are racist for preferring tall candidates to short ones.

Posted by: Anarchus on December 13, 2002 07:32 AM

what percentage of American Jews are observant Orthodox? Surely the odds of Sophie Schwartz (who may be Greek married to a Jew for all our HRC manager knows...

For the purpose of my illustration, I made it a point to reference an address in a neighborhood famously heavily populated by Orthodox Jews.

BTW, I've employed an orthodox woman who was one of my best-ever workers. We mapped out her work year well in advance, & the trade-off worked as follows: she got all her holidays off, while I got someone who was loyal and diligent. My business flexibility allowed for that.

Posted by: George Zachar on December 13, 2002 07:37 AM

question the work ethic of Jews (sorry, George)

No need to apologize. You're the one who misread my post and turned it into an insult.

Posted by: George Zachar on December 13, 2002 07:48 AM

David,
OK, you support good programs. Which ones? Give me a fer instance...

Posted by: Emma on December 13, 2002 08:26 AM

As long as we're going around the bend here, accusing folks of being racist and anti-semitic, I thought I'd toss this molotov cocktail into the mix:

There is a good chance that an odd cluster of hereditary neurological diseases among the Ashkenazi Jews is a side-effect of strong selection for increased intelligence.

http://www.jerrypournelle.com/reports/cochran/overclocking.html

Posted by: George Zachar on December 13, 2002 08:54 AM

All in favor of deporting Thomson to Israel where he would have to beg for Christian holidays off (actually, his right to take them off is unconditionally protected by Israeli law), say "Aye". Can anyone figure out where he works and drop a dime to the EEOC, newly-reminded that bigotry really does exist? It says wonders that he ignored the fact that the resumes were almost certainly set up to control for college GPA and major (I always look at major when reviewing resumes).

How much would Mr. Thomson like to bet that employers were less likely to call back black applicants long before Affirmative Action? Check the novel (and award-winning movie) "Gentleman's Agreement" for anti-Jewish discrimination at a time that a Jew's diploma probably meant more than a non-Jew's because there was also discrimination in college admissions.

What excuse is next, David? Sunspots?

Discrimination is out there. Anyone remember the great Ian Ayres studies that car salesman consistently offer lower prices to whites than blacks?

Posted by: Andrew Lazarus on December 13, 2002 08:57 AM

Yeh. Right. "Untraditional names". That's what people are really prejudiced against.

woteva.

dd

Posted by: dsquared on December 13, 2002 09:12 AM

“I'm still mystified. Our very own Trent Lott, David Thomson, says this;

"Unfortunately, it is fair to wonder if a black person's degree is worth the paper it is printed on."”

Why is so hard to understand that this would be the inevitable results when the Liberals dropped the academic standards? Why blame me for pointing out reality? Believe or not, this is not my fault.

“All in favor of deporting Thomson to Israel where he would have to beg for Christian holidays off (actually, his right to take them off is unconditionally protected by Israeli law), say "Aye". “

What have I said to even begin justifying such a response? Oh well, I guess I will just shrug this off.

“How much would Mr. Thomson like to bet that employers were less likely to call back black applicants long before Affirmative Action? “

Do you have a problem reading English? I definitely believe that affirmative action programs are justified to address past abuses. How much clearer do I have to be?

“David,
OK, you support good programs. Which ones? Give me a fer instance...”

I cannot point out to any particular programs. After all, I was not involved in their creation. I have, though, repeatedly said that sometimes it is legitimate to hire the minority candidate---if everything else is equal! Unfortunately, often a grossly inferior minority applicant has been chosen over one more qualified.

Posted by: David Thomson on December 13, 2002 09:45 AM

>>Patrick, how certain can you be that you do not have some unconscious bias in your process that makes you less likely to select equally high productivity/wage black candidates than white candidates? There are several steps in the process you describe that are potentially subject to unintentional bias.<<

All human beings are subject to unintentional bias, so the question is one of degree. To which I point again to the difficulty of finding productive (people who will help me make money) employees. There's a strong incentive for me to get the right people.

BTW, I know other entrepreneurs who follow the same hiring practice, some of whom are actively hostile to college graduates. The reasoning being that anyone so stupid to waste four years to gain a virtually worthless credential (and to go into debt to boot)is also too stupid to be a productive worker.

Posted by: Patrick R. Sullivan on December 13, 2002 10:47 AM

>> 'To make the obvious point, black college graduates are much more likely to have been admitted under lower academic standards/test scores, and employers know this.'

>> What they have to do to graduate is the same as white students. This is also known to employers.<<

Does Jason McCullough "know" this firsthand, as an employer with his own money at risk?

Posted by: Patrick R. Sullivan on December 13, 2002 10:57 AM

"I'm not going to call this girl, I don't even know how to pronounce her name. I'd make a fool of myself."

I think a lot of this comes down to a bias against strange sounding names, not because of a perceived skin color based on name.

Posted by: Unseelie on December 13, 2002 10:58 AM

Ummm ... what proportion of black employees actually file racial-discrimination complaints? How does an employer's risk that a black employee will file an unjustified discrimination complaint measure up against all the other risks that an employee will turn out to be incompetent?

And as long as an Orthodox Jewish employee works late on Monday-Thursday, and maybe comes in on the occasional Sunday, to make up for leaving early on Friday, where's the problem? (OK, I can see how certain jobs require someone's physical presence in the office from 3-5 pm, but as long as the employer makes that expectation clear, the Jewish candidate would either not take the job or would have no cause to complain about the hours.)

Posted by: Seth Gordon on December 13, 2002 11:19 AM

It also looks ridiculous on its face, even before you consider the credence its given in anti-semitic circles.

'>> What they have to do to graduate is the same as white students. This is also known to employers.<<

Does Jason McCullough "know" this firsthand, as an employer with his own money at risk?'

You're right, Patrick: perhap employers, in general, are horribly misinformed about what you have to do to get a college degree. God knows that's a much more reasonable explanation.

Posted by: Jason McCullough on December 13, 2002 11:20 AM

It also looks ridiculous on its face

I thought so too, but Jerry Pournelle is a bright guy who doesn't put troll food up on his site.

The story's core idea, that certain genetic and demographic issues correlate with tested IQ, is thought provoking, whether or not one likes its impact.

Posted by: George Zachar on December 13, 2002 11:30 AM

It's also highly unlikely that a detectable proportion of the hiring bias seen is based on "the risk of increased discrimination claims." As the EEOC's statistics point out, there's only about 30,000 racial discrimination charges filed with them a year; 2/3rds of those are dismissed. With 16 million african-americans in the labor force, that's a claim rate of 1 in 533, and a compensated claim rate of 1 in 1600.

Maybe there's a hidden explosion of local and state level claims, but I can't find the relevant statistics on the web. Unless some truly large numbers from them are presented, there's no way discrimination claims are an explanation for the results of this study.

Posted by: Jason McCullough on December 13, 2002 11:39 AM

He may be a bright guy, but it also looks like Pournelle's a closet Bell Curver:

'I think there is no serious (other than politically correct) disagreement among informed experts that IQ has a strong hereditary component. Murray and Herrnstein put it at 60%; that's actually considered low.'

Ugh.

Posted by: Jason McCullough on December 13, 2002 11:43 AM

what proportion of black employees actually file racial-discrimination complaints?

A far larger proportion than that of white employees, making such litigation, net, an additional employment risk for one group of candidates

This is a textbook classic example of the law of unintended consequences.

Posted by: Bucky Dent on December 13, 2002 11:45 AM

Murray and Herrnstein put it at 60%

Hmmm. I recall them putting it closer to 50%.

Posted by: George Zachar on December 13, 2002 11:49 AM

Folks, there is a possibly too subtle difference between disagreeing with the methodology of this study, and disagreeing with its conclusions. It may well be that critics of this study also believe that there is racism in the workplace, but don't consider this study to be convincing proof of it.

Anyway, as to the question of how likely a black person is to file a wrongful dismissal or workplace discrimination lawsuit with the EEOC - the premise of the article is that the resumes are "equal". If this legal risk is not equal, even a slight edge to the white applicant could produce a big swing in results.

So I suppose the real question is, might the sense that such a legal threat exists reasonably be in people minds? One highly publicized lawsuit could have a much greater public relations impact then a lot of quietly settled lawsuits. Still, I imagine that proportionately more blacks than whites file race-based discrimination lawsuits. Haven't checked, though, but the questions seems to be sufficiently reasonable that the authors might have addressed it in their study.

Similarly, the statement that the college admissions process is influenced by racial preferences seems to be generally accepted. Now, as a casual reader of the newspapers, I know that some municipalities practice "race based" norming on things like civil service exams for municipal hiring. E.g., the top ten percent of white applicants get jobs, and the top ten percent of black applicants get jobs. This can get controversial when a black person is hired despite having a lower score than a white who was passed over.

Now, could such a thing happen in college? If, in four years, a black student has one professor who grades this way, it is possible (depending on how all the students perform in the class) that the black student's grade in that class, and hence his overall GPA, got a boost.

If this is a possibility, than it suggests that, when comparing GPAs and majors between whites and blacks, "equal" isn't equal. Now, I think D Thompson, former NBA All-Star, goes a bit overboard in saying that the degree for a black person may not be worth the paper it is printed on. But if folks can reasonably believe that this sort of preferential grade inflation occurs, then the resumes will not be considered equal, and the study has a problem.

Now, Jason asserts (on no evidence) that no reasonable employer would believe this scenario to be possible. Well, as a casual observer of the college scene, I would not be surprised to learn that it happens, and would not be surprised to learn that people suspect that it might happen. Given the obvious preferential treatment on the admissions side, I would think the authors of this study would want to offer evidence that prospective employers do not contemplate preferential grading in evaluating GPAs.

And to repeat - if all these objections are valid, it doen not follow that Affirmative Action is bad, or that discrimination does not occur. It just means this study has problems.

Posted by: Tom Maguire on December 13, 2002 12:12 PM

In the early 90's I worked in a copy shop in Brooklyn. We recieved a resume from a black woman with the last name of 'Cohen.' I asked her if she was married to a Jewish man, she said no, she had changed her last name to Cohen. She said that in New York City it made everything go so much more smoothly on the phone.

Posted by: Bruce Lokeinsky on December 13, 2002 12:12 PM

Poor "Keisha" (Table 2, p. 26 in the study)got only 3.8% callback (3rd worst out of 26 names. Can anyone imagine why she would so badly underperform "Ebony" (10.5%, and 8th best out of 26)?

Especially since some Keisha's look to be race neutral:

http://www.nba.com/pacers/dance/keisha.html

I also see another obvious problem with this study, it was conducted between July 2001 and January 2002. Meaning more than half of it came after the WTC attack, which is not a typical labor market. Further, of the "black" male names, Rasheed, Kareem, and Hakim, would have Arab connotations.

Posted by: Patrick R. Sullivan on December 13, 2002 01:01 PM

Jason wrote (quoting Pournelle):

'I think there is no serious (other than politically correct) disagreement among informed experts that IQ has a strong hereditary component. Murray and Herrnstein put it at 60%; that's actually considered low.'

Ugh.
----

Yep. Jerry must have missed the study (published in Nature, can't recall the authors) on fraternal versus identical-twins. Turned out that sharing an interuterine environment accounted for ~30% of variation, which pretty much totaled M & H's hypotheses of a genetic cognitive elite.

He's also not distinguishing between weak-sense and strong-sense heriditability (but, to be honest, I have a hard time remembering the distinction meself).

Also, remember that hereditability is also a factor of the amount of variability in the environment; so a high degree of hereditability does not imply that environment is not a factor.

Posted by: Tom on December 13, 2002 01:18 PM

>> Interviews were requested for 10.1 percent of applicants with white-sounding names and only 6.7 percent of those with black-sounding names....<<

IOW, racially neutral names such as Emily and Jay (I know blacks of said names)did not get callbacks 90% of the time. While for "black-sounding" names the figure was 93%. In TWO cities (no doubt both carried by Al Gore in 2000), during a calamitous business climate.

And, "Tamika" (94.6% non callback rate, 8th worst of 26 names), I would associate with Japanese, not African-American.

Posted by: Patrick R. Sullivan on December 13, 2002 01:27 PM

“ As the EEOC's statistics point out, there's only about 30,000 racial discrimination charges filed with them a year; 2/3rds of those are dismissed. With 16 million african-americans in the labor force, that's a claim rate of 1 in 533, and a compensated claim rate of 1 in 1600.”

What world do you live in? The mere fact that an employee files a discrimination suit against a small company is enough to scare the hell out of them. So what if 2/3rds are later dismissed? The financial and time costs can be enormous. My guess is that the process might cost them a minimum of $5,000 and this a rough hit for owners who are lucky to declare annual gross earnings of$60,000. Please take me to task if you think my numbers are out of line. However, I suspect that I’m at least in the ballpark.

Posted by: David Thomson on December 13, 2002 01:27 PM

Re:

>>IQ has a strong hereditary component. Murray and Herrnstein put it at 60%.<<

Ummm...

First of all, yes, IQ does have a strong hereditary component. I have a much higher IQ than the banana on the table to my left, and the overwhelming component of that IQ difference is due to my superior intelligence genes.

Or, for another example, consider two populations which share identical genes, but one (the "serfs") are routinely massively protein-deprived before birth and during childhood, and the other (the "lords") are well-nourished throughout their development. All the difference between the average intelligence of the two groups is due to environment, yes?

The "60%" number (which I don't trust, because I don't trust Charles Murray as far as I can throw him to perform *any* calculation accurately) depends on a *specific* pattern of environmental differences, and a *specific* pattern of genetic variation--double the variation in the environment, and you push the "hereditary" component down to 43%; quadruple the variation in the environment, and you push the "hereditary" component down to 27%...

Posted by: Brad DeLong on December 13, 2002 01:31 PM

“ As the EEOC's statistics point out, there's only about 30,000 racial discrimination charges filed with them a year; 2/3rds of those are dismissed. With 16 million african-americans in the labor force, that's a claim rate of 1 in 533, and a compensated claim rate of 1 in 1600.”

What world do you live in? The mere fact that an employee files a discrimination suit against a small company is enough to scare the hell out of them. So what if 2/3rds are later dismissed? The financial and time costs can be enormous. My guess is that the process might cost them a minimum of $5,000 and this a rough hit for owners who are lucky to declare annual gross earnings of$60,000. Please take me to task if you think my numbers are out of line. However, I suspect that I’m at least in the ballpark.

Posted by: David Thomson on December 13, 2002 01:32 PM

Jason M: I don't know what Pournelle's views are on this issue, but there's nothing 'Bell Curvish' about the quote you provide. Performance on IQ tests (and what those actually measure is a whole other can of worms) does indeed have a moderately strong genetic component. This comes from studies of identical twins raised in different homes and adoption studies. It's an orthogonal issue to cross-racial differences in IQ though, because it in no way denies that your environment also plays a very large role.

And that's where the argument against the Bell Curve comes in. Those authors claimed to have controlled for all environmental variables, and still found a leftover 'purely' genetic difference between blacks, whites, and asians. Most people don't buy that claim though, and still ascribe the differences they found to environmental factors.

Anarchus: Maybe you missed my post above. The "name" issue probably did contribute some effect to the results, no argument there at all, but the way they designed their study worked around that problem.

You might expect to find that, across the board, "unusual" names got fewer callbacks than "usual" names. However, they then had a second manipulation that looked at the effect WITHIN each name group of having more experience.

It's reasonable to suspect that, overall, the unusual names would get fewer call backs (just because of boring sorts of "familiarity" effects, or something like that). It isn't reasonable though to expect that, within the subgroup of unusual names, level of experience wouldn't have any effect.

In other words, if the "name" factor was driving the results, you would have lower rates of callback overall for unusual names, but the same PATTERN of high vs. low experience within each name group.

But that isn't what they got. They got lower overall rates for black names, AND a flat effect of experience only for black names.

I'd also like to back Anarchus up on the legitimacy of his question. There have been some vile comments thrown around on this thread, but Anarchus's wasn't one of them. Just because you agree with the outcome of a study doesn't mean the methodology gets a pass (in fact, that's exactly why politically charged issues sometimes suffer from lackluster research, because people are too emotionally connected to the outcome.
It's much cleaner to study non-loaded topics such as the mating habits of horshoe crabs :) )

Anarchus asked a valid (and interesting) methodological question. The study appears to have adequately addressed it though.

Posted by: rufus on December 13, 2002 01:46 PM

"Surely the odds of Sophie Schwartz (who may be Greek married to a Jew for all our HRC manager knows) taking a lot of days off are not much different from Jane Smith being a sickly."

Yes, but what if Sophie Schwartz is an Orthodox Jew who is also sickly? Or worse yet, an Orthodox Jew who is sickly, and files a lot of religious and medical discrimination suits? ;-)

Well, if Sophie is drop-dead gorgeous, all that won't matter! She'll be the one that's hired! ;-) (For details, see my comment of 4/12, 4:33 pm.)

Posted by: Mark Bahner on December 13, 2002 02:10 PM

“First of all, yes, IQ does have a strong hereditary component. I have a much higher IQ than the banana on the table to my left, and the overwhelming component of that IQ difference is due to my superior intelligence genes.”

Since when does a banana have an IQ? Oh well, I will let that pass and share my own concerns about the Bell Curve. The authors should have stressed that human intelligence is a forevermore intrinsically nebulous area of study. Studies currently released will likely be contradicted in the future. The nature/nurture debate will never be totally resolved. Therefore, it’s best to take a chill pill when any work on intelligence is released. Humility is mandatory because the odds are high that someday people will be laughing at your tentative conclusions.

Posted by: David Thomson on December 13, 2002 02:16 PM

>> Interviews were requested for 10.1 percent of applicants with white-sounding names and only 6.7 percent of those with black-sounding names... <<

IOW, resumes with racially neutral names such as Emily or Jay, were not contacted 90% of the time. While "black-sounding" names were not contacted 93% of the time.

In two cities carried by Al Gore in the 2000 election. During a dramatically atypical labor market.

Posted by: Patrick R. Sullivan on December 13, 2002 02:26 PM

'Now, Jason asserts (on no evidence) that no reasonable employer would believe this scenario to be possible. Well, as a casual observer of the college scene, I would not be surprised to learn that it happens, and would not be surprised to learn that people suspect that it might happen. Given the obvious preferential treatment on the admissions side, I would think the authors of this study would want to offer evidence that prospective employers do not contemplate preferential grading in evaluating GPAs.'

Let me state this clearer: unless evidence is presented supporting the assertion that minorities receive "preferential grading" as university students, it is a disturbing, if not racist, opinion. "I wouldn't be surprised if it happens" doesn't cut it on a radioactive subject like this.

They state on page 18 that "only about 11% of the ads list an explicit education requirement. Of these, 8.8% require a high-school degree, 49% some college, and the rest a 4-year college degree."

As to the potential lawsuit risk for African-American hiring: let's do some calculations. Since no one is going to defend risk-aversion from cases where the employer actually is racist (the 33% that aren't thrown out), let's just use the cases that aren't considered actionable by the EEOC (1 in 800) as a proxy for bogus discrimination claims.

The cost of those bogus claims increases the expected cost to employers when they hire an african-american by 1/800 * salary * cost of claim.

Example % increases in employer labor cost for various costs of bogus claims:

1/3 * salary = .04167% increase.
1 * salary = .125% increase.
8 * salary = 1% increase.

It's pretty hard to to justify the results of the study with the above numbers. Even if you throw in claims where the employer is at fault, the increased cost is amazingly small.

Posted by: Jason McCullough on December 13, 2002 02:26 PM

'Jason M: I don't know what Pournelle's views are on this issue, but there's nothing 'Bell Curvish' about the quote you provide. Performance on IQ tests (and what those actually measure is a whole other can of worms) does indeed have a moderately strong genetic component. This comes from studies of identical twins raised in different homes and adoption studies. It's an orthogonal issue to cross-racial differences in IQ though, because it in no way denies that your environment also plays a very large role.'

He cites Murray and Hernstein's estimate like it's a real number, when everything from that intellectually dishonest book should be held with iron tongs. That's enough for me.

Posted by: Jason McCullough on December 13, 2002 02:37 PM

Fair enough.

It sounded as if you were decrying the mere existence of any genetic component at all. If your problem was merely with the particular percentage cited, we'd be in agreement.

Posted by: rufus on December 13, 2002 02:55 PM

Well if I wasn't convinced that Jason McCullough has zero business experience before, after his latest I'm 100% sure of it:

>> let's just use the cases that aren't considered actionable by the EEOC (1 in 800) as a proxy for bogus discrimination claims.

>> The cost of those bogus claims increases the expected cost to employers when they hire an african-american by 1/800 * salary * cost of claim.<<

So, Jason considers the EEOC to be a disinterested bystyander? Jason also thinks legal advice for people accused of racial discrimination is free?

And perhaps the opinion of an expert is in order:

"Civil rights laws were not passed to protect the rights of white men and do not apply to them."
-- Mary Frances Berry, Chairwoman, US Commission on Civil Rights

Posted by: Patrick R. Sullivan on December 13, 2002 03:09 PM

He cites Murray and Hernstein's estimate like it's a real number, when everything from that intellectually dishonest book should be held with iron tongs. That's enough for me.

Just curious. Have you actually read The Bell Curve?

Posted by: Bucky Dent on December 13, 2002 03:25 PM

'Jason also thinks legal advice for people accused of racial discrimination is free?'

Uh, so you missed the sentence where I estimate the legal costs of false discrimination claims? You know, one of the two you quoted?

'Just curious. Have you actually read The Bell Curve?'

I was a bit young when it came out, so no. Of course, this is a wonderful springboard for you to accuse me of not knowing what I'm talking about, blah blah blah, as if the book's statistics have been somehow misrepresentated.

'If your problem was merely with the particular percentage cited, we'd be in agreement.'

It's more that the entire book is shoddy, and if you can't recognize that you're a bad source. Which is Pournelle's problem.

Posted by: Jason McCullough on December 13, 2002 03:57 PM

'Just curious. Have you actually read The Bell Curve?'

I was a bit young when it came out, so no.

You might consider actually reading the book before condemning it in vivid prose.

Posted by: Bucky Dent on December 13, 2002 04:25 PM

Tom Maguire says:

"The authors seem to be assuming that the correct callback ratio for interviews should be 50/50. Fewer blacks being called back is evidence of racism.

Well, it may be evidence of non-colorblindness in the hiring process. However, my (laymens) understanding of EEOC guidelines is that a company's workforce should reflect the ethnic composition of the region. So, suppose Massachusetts has two whites for each black. A company might be calling back and hiring two whites for each black - no racism at all, EEOC approved! And if all companies did this, we would have an environment with no overt hiring racism, but hideous name-oriented racism based on studies like this.

As I said, I don't know how the authors controlled for this."

Wow. This thread has been an amazing lesson in how little prodding is required for conservatives to show their true colors. Let's take this remarkable piece of analysis by Tom Maguire. If the fraction of blacks in a region is 'x', then all things being equal we would expect the fraction of applicants to be 'x'. Then to have a workforce with a fraction 'x' of blacks, the blacks would need to be hired at an equal rate as non-blacks, no? x*p with p<1 does not equal x, right?

It's amazing what someone will find to convince themselves of what they want to be convinced of. Kinda depressing.

Posted by: on December 13, 2002 04:35 PM

Damn, did it again. The post above starting with "Tom Maguire says:" was mine.

Posted by: Ben Vollmayr-Lee on December 13, 2002 04:37 PM

'You might consider actually reading the book before condemning it in vivid prose.'

Apparently you ignored the rest of my post. Here's a deal: I'll read the Bell Curve if you slog through the collected works of Noam Chomsky. We both lose!

Posted by: Jason McCullough on December 13, 2002 04:47 PM

I've read Chomsky, thank you. Linguistics and politics.

Posted by: Bucky Dent on December 13, 2002 05:02 PM

Mr. Rufus: I appreciate the support. There are two issues here: (1) The study, and (2) Racial prejudice in America. I've not commented on (2) and don't intend to start now.

I still maintain the study is flawed. While the authors examine the intra-race name statistics, as best I can tell pretty much all of the African American names (Ebony, Aisha, etc) are very non-standard, and all of the White names are very standard. And I understand that they went by the SSN data to ground the analysis in real names, but the fact remains that there's a very strong underlying trend for African American parents to use modern names while White parents for the most part have stuck with more traditional names. I still think that the authors did not control for this variable properly. I may be overly sensitive to the issue, but my wife and I have had four children in the last seven years and I feel like I know the SSN name data by heart at this point - especially because my wife was very biased against "common" names, so we had to search extra hard for names that weren't ranked in the top 50 in our state (PA). For whatever it's worth, our four names are: Lucy, Clara, Sophie and Cecilia. As best I can tell, they are NOT in the study.

Anyway, I'm getting fed up with the name-calling and racial overtones. Bye.

Posted by: Anarchus on December 13, 2002 05:36 PM

If you haven't read it, how do you know it's shoddy?

Posted by: tc on December 13, 2002 06:36 PM

Lucky me, I dont' have to read it.

'If you haven't read it, how do you know it's shoddy?'

If someone tells you that a math book insists 1+1=3, and you verify that it does, do *you* read it?

Posted by: Jason McCullough on December 13, 2002 07:28 PM

Each of these people are individuals and, as such, you have no idea where they fall on the Bell curve, even if you accept it's ugly premise.

Bob Smith who went to Northwestern might have been a linebacker "business major" who had a lot of help from the athletic department while Taisha Jones might have been a top scholorship student in the math department at the same school.

(And for all you know, Taisha was named after her mother's college roommate and Bob is known as MC Cracked Ice to his old friends from South Central. To make any assumptions about their race or their religion and the supposed "risks" those identifications present based upon their name or their neighborhood is absurd.)

But, even if you did know their race or religion, making judgements about individuals based upon statistical probabilities (or more likely your own flawed observations) that they share "risky" characteristics with others in their "group" is the definition of prejudice.

Because, even if we accept that 60% of blacks score lower on IQ tests than whites, what exactly does that tell you about Taisha? Taisha could be a genius, for all you know, smarter than anyone else in the pile of resumes but because you make negative assumptions about people based upon their race, EVEN IF THEY HAVE THE SAME QUALIFICATIONS AS THOSE YOU DECREE LESS RISKY, you will miss out on the better worker.

And what kind of system says that it is moral or intelligent or rational that Taisha, a top student and hard worker, should be discriminated against because you believe that since 60% of her racial group score lower on IQ tests, or because a higher percentage of her race file discrimination suits, that as far as you're concerned it might as well be 100%?

And what about Bob Smith? Is it any more reasonable to assume that Bob scores higher than Taisha because he's white? He could just as easily be in the lower percentile of ALL racial groups.

When you make these kinds of assumptions about individuals you are effectively treating all of those who are in your "risky" group as if they are at the bottom of the curve and giving all of those in your "non-risky" group the benefit of the doubt that they fall into the upper. You might be happy enough with Bob, all things being equal. But, you have been tremendously unfair to Taisha and that is shameful.

Most people aren't Taisha, to be sure. But, a fair number are. And it is a terrible thing to make them struggle so much harder to even get in the door just because you can't be bothered to discard your assumptions and treat everyone as an individual.

To believe that you are making a rational hiring decision about two individuals based upon these broad statistical probabilities is fooling yourself. You are simply a bigot.

Posted by: digby on December 13, 2002 08:07 PM

Well, Anarchus has a point - being called a racist for having questions about a study does get a bit tiresome.

As to Jason's point, Let me state this clearer: unless evidence is presented supporting the assertion that minorities receive "preferential grading" as university students, it is a disturbing, if not racist, opinion. "I wouldn't be surprised if it happens" doesn't cut it on a radioactive subject like this.

Well, does this mean you are not open to discussing the possibility of preferential grading occurring? This is only radioactive if we decide that it is - I suppose a civil discussion is an alternative. And my point was that such a view might prevail in the hiring community at large - your view is that any such thought is so unreasonable as to be unworthy of discussion, and bordering on racist?

I am not expressing an opinion about the capabilities of black or white students - I am commenting on the various pressures and motivations that might affect the faculty. If I am current on my reading, many faculty members "tilt left", and at least some might choose to grade black students more sympathetically, maybe. Other profs might want to ward off any possible complaints from students by handing out higher grades to students likely to be able to sustain a complaint. These are things we cannot discuss?

And the helpful Ben V-M has this:

Wow. This thread has been an amazing lesson in how little prodding is required for conservatives to show their true colors. Let's take this remarkable piece of analysis by Tom Maguire.

Analysis? I asked a question about how the authors controlled for the likelihood that blacks and whites are not 50/50 in the workforce. Evidence of my uncertainty can be found in the phrases so whoever read the actual study may be able to help me here... and
...As I said, I don't know how the authors controlled for this.

Anyway, Ben, are you telling us that the authors sent out applications in the same proportion as workers in the workforce to control for this? or that it doesn't matter? And are you able to separate the analysis from the invective, or is it a package deal? Based on your conclusion It's amazing what someone will find to convince themselves of what they want to be convinced of. Kinda depressing, I would hate to deoress you further.

You also may have missed my comment on another post, so I'll repeat it. I'll even go with bold tags, just to help.

And to repeat - if all these objections are valid, it doen not follow that Affirmative Action is bad, or that discrimination does not occur. It just means this study has problems.

Posted by: Tom Maguire on December 13, 2002 08:49 PM

Just to depress Ben further, I will illustrate my point about the possible significance of workforce ratios with some examples.

Suppose a small company normally hires two whites for every black, as per local workforce ratio. By happy coincidence, it has received three "genuine" applications for three job openings, two from whites, one from a black. At this point, everyone is on track for a callback. Now, two more applications arrive in the mail, one from a white, one from a black. Both applications are "fake", being from the researchers.

So, if the company still wants to call back two whites, there is a 67% chance of the researchers getting a call-back (assuming, for no good reason, that the "fake" applicant is equal to the two real ones. Or, assuming that the fake will vary unpredictably relative to the real ones).

However, for the black "fake", the chance of a call back is 50%. In this primitive example, the white applicant recives more callbacks, even though the company is not (I don't think) biased.

Now, how about at a larger company - suppose they have 66 real white applications, and 33 real black applications, still with a 2-1 target callback ratio. Add one fake application to each color. Now, the chance of a callback for the white is 2/67, versus 1/34 for the black. These ratios are essentially equal.

Hmm. So if the applications are going in to large companies, then the ratio being mailed out by the researchers doesn't matter - it gets buried in the much larger pool.

OTOH, the ratio used by the researchers does seem to matter when the pool of "real" applicants is small.

Anyway, those are pretty much my true colors on that point. As I said, I am going by the newspaper summary, so I don't know if the authors addressed this. It popped into my head as worth asking about.

Posted by: Tom Maguire on December 13, 2002 09:14 PM

To Tom Maguire: the problem with the idea that professors would grade certain groups of students differently based on race is that in any class larger than a seminar, the professor often does not know the race of his/her students. They see the faces every day, to be sure, but unless they actually have some sort of student-teacher relationship (which is virtually non-existent in lecture classes going by my experience thus far), they don't connect faces with names. It seems to me that your argument depends on this ability to match (black) faces with the names on a paper or an exam, and I don't see how that could happen in the lecture classes which make up the majority of college classes.

Posted by: Mark on December 13, 2002 09:44 PM

To Tom Maguire: the problem with the idea that professors would grade certain groups of students differently based on race is that in any class larger than a seminar, the professor often does not know the race of his/her students. They see the faces every day, to be sure, but unless they actually have some sort of student-teacher relationship (which is virtually non-existent in lecture classes going by my experience thus far), they don't connect faces with names. It seems to me that your argument depends on this ability to match (black) faces with the names on a paper or an exam, and I don't see how that could happen in the lecture classes which make up the majority of college classes.

Posted by: Mark on December 13, 2002 09:45 PM

Tom, going by the newspaper summary, they fake applicants had an interview request rate of about 8.4% total. We can get an estimate of what fraction the fake candidates contributed to the applicant pool if we assume their resumes were typical (also stated in the article): about 12 to 1. An average of 4 fake resumes were sent in for a job advertisement, so these ads were averaging about 50 applicants, give or take 25. These numbers are all in the article and digested pretty intuitively at read-time by anyone with analytic reasoning skills, which I would assume to include you. Hence, it looks to me like you're fishing for a way to avoid accepting something you don't like to hear.

Posted by: Ben Vollmayr-Lee on December 13, 2002 10:00 PM

Ugh. Analytic reasoning skill clearly doesn't indicate editing skill. Strike the 2nd sentence above: I meant to say 1/0.0084 ~ 12, so real applicant to fake applicant ratio was about 11:1. The rest follows coherently.

Posted by: Ben Vollmayr-Lee on December 13, 2002 10:06 PM

Meant 1/0.084, obviously. Maybe I should use the "preview" feature!

Posted by: Ben Vollmayr-Lee on December 13, 2002 10:10 PM

Hmmm, it might not be troubling that you haven't read 'the Bell Curve' Jason, but it also sounds as if you're not very clear on what it is about the book that is controversial.

Apart from the claims about cross racial differences, the rest of it is basically benign academic stuff that reasonable people can disagree about (e.g. the % hereditability figure which has been discussed here). It's really only their position of ascribing racial test differences purely to genetics, rather than environment, that is viewed as beyond the pale.

One can quite easily believe in a very strong genetic component to intelligence (higher than even the BC authors claim) and STILL believe that inter-racial differences are entirely due to environment. The 60% figure is likely too high (imo), but there's nothing inherently dishonest or racist about it on it's own.

Anarchus: it seems as if we're talking past each other here a bit, so I won't belabor things further.

Posted by: rufus on December 14, 2002 12:31 AM

If racism is not an issue in the US how come this comments thread is so much longer than any other of Prof DeLong's? There have been some amazing attempt to deny the obvious here.

I actualy think everyone so far on this thread has confused social wellbeing with individual rationality (the libertarians are anyway prone to deny a priori that the two can differ). 'Statistical discrimination' (discriminating in favour of someone because the group they're in has a disproportionately higher incidence of some desired quality, and you have no costless way of judging that person's individual qualities) can be perfectly rational for an individual - it's a consequence of asymmetric information.

But statistical discrimination is just as unjust as true bigotry, and is harder to eliminate precisely because it aligns with the decision maker's rational self interest. It also generates economic inefficiencies - it can trigger a self-reinforcing process, where the privileged gather ever more of these desirable qualities (education, good work habits, staying within the system etc) because they get bigger payoffs from them, and so are given ever more privileges. The potential of those excluded is never realised.

You can escape this local optimum to reach the global optimum (where privilege is not decided by arbitrary group membership) by enforcing a countervailing positive discrimination.

There is an old literature on economic theories of discrimination that formalises all this. But while we assume that all discrimination is individually irrational we will never be able to counter its effects (however much ranting and raving against 'bigots' givues us a warm inner glow). And if we identify individual rationality with social optimality we'll be constantly tempted to deny discrimination's existence.

Posted by: derrida_derider on December 14, 2002 04:30 AM

"If someone tells you that a math book insists 1+1=3, and you verify that it does, do *you* read it?"

I would first ask if this is a printing error. Also, many books are weak in some areas and not in others. One of the important reasons for the Bell Curve's difficulties is probably due to the health problems of one of its authors. Richard Herrnstein was so ill that he died about the time the book was released.

Posted by: David Thomson on December 14, 2002 04:31 AM

Mark: I accept your point about lecture classes. However, mypoiny is not that all,or even most classes might grade preferentially. If an employer thinks even one grdae over four years is inflated, the employer might mentally adjust the black applicants GPA downward by .01, from example, from 3.0 to 2.99. At that point, the white and black applicant's do not have resumes thatare perceived as equal. If the employers mental adjustment reflects his guess as to what truly happens in the course of four years at college, it is also not racist, IMHO. It is simply discounting for the race based behavior of a few professors.

Some evidence that something like that goes on is offered in the newspaper article here:

Their most alarming finding is that the likelihood of being called for an interview rises sharply with an applicant's credentials — like experience and honors — for those with white-sounding names, but much less for those with black-sounding names.

My guess would be that folks, evidently,don't take the awrds as seriously when given to a black person. I continue to believe that this does not reflect on the black person themselves, but simply on the award-giving environment.

For a specific hypothetical example, consider two applicant who put five years in a civil service job on their resume. I can not speak to Chicago and Massachusetts (which would seem to be relevant, but it is my example), but many civil service departments have entrance tests and race based norming (I think I am repeating myself, sorry). The consequebce is that, on occassion at least, the average black "new hire" has a lower civil service test score than the average white "new hire". The same thing can happen at vaious promotion points, I believe.

So, we have two resumes with seemingly equal work experience in the civil service. Which one does a prospective employer give more weight to?

Which does not prove Affirmative Action to be bad - it simply means that, along with whatever good it is doing, Af. Ac. will have "bad" side effects, for example in reducing the signalling effect asociated with past achievement.

Now, Ben: OK, the average applicant pool may have been fifty. In a large company, as I pointed out in my example, the workforce ratio effectdisappears. In the small company, it lingers. Maybe every test was conducted at an "average" company. Oherwise, they probably sent resumes to some mix of larger and smaller companies, with some effect on the results. Probably a slight effect, which is why I wondered whether the authors had identified this and thought it needed to be addressed.

As to whether I am fishing for reasons to object to this study, your proposed alternative is what? Blind acceptance? Eventually, a study may come along with conclusions to which you object - what then?

Posted by: Tom Maguire on December 14, 2002 04:34 AM

"...the problem with the idea that professors would grade certain groups of students differently based on race is that in any class larger than a seminar, the professor often does not know the race of his/her students."

Do you think that nobody reads the newspapers, or watches the evening news? I’m sorry but only a hermit living on an isolated island doesn’t know that the standards have often been lowered for black students. Even Rice University in Houston has finally surrendered its integrity. Columbia and Harvard sold their souls to the devil years ago. The latter is infamous for its grade inflation scandals. It is also common knowledge that all minorities are not treated the same. Asians, for instance, are required to have the same test results as whites.

Posted by: David Thomson on December 14, 2002 04:56 AM

“And I want to thank my father right now for giving me the particularly white name of Wesley.

Thank you, thank you, thank you.
Posted by at December 12, 2002 01:41 PM”

The very first post on this thread inadvertently highlights the inherent weakness of this study. Has anybody ever heard of Wesley Person? If not, you might like to take a look at the following link:

http://sports.yahoo.com/nba/players/2/2646/

Posted by: David Thomson on December 14, 2002 05:03 AM

"'Statistical discrimination' (discriminating in favour of someone because the group they're in has a disproportionately higher incidence of some desired quality, and you have no costless way of judging that person's i
ndividual qualities) can be perfectly rational for an individual - it's a consequence of asymmetric information."

Yes, if each buyer were acting in isolation. But if other people are practicing statistical discrimination, the payoff to you of gathering more information rises (since your competitors have picked talented individuals from the statistically favored group, but not from the statistically disfavored group, your chances of finding your talent in the statistically favored group falls faster than your chances of finding talent in the statistically disfavored group).

This is probably the mechanism by which discrimination against other formerly disfavored groups (Irish, Asians, etc) disappeared. You will note that discrimination against these groups were not stamped out by government fiat, so there must be another explanation, and this seems most likely to me.

Posted by: Kenneth Uildriks on December 14, 2002 05:23 AM

To Tom Maguire: You're still assuming that some professors do grade on that sort of a basis ("It is simply discounting for the race based behavior of a few professors.") without any evidence. How would race-based grading take place? Do you have any evidence that it in fact does take place? Speaking as a college student at a place where affirmative action certainly does exist, I've never heard of such a thing, and I would be very much surprised if it went on at my school. But let's say it does on a small basis, and a firm gets a resume from a black graduate with a couple of grades inflated along those lines. Assuming they believe the kind of grading we're talking about takes place, how are they do distinguish the inflated grades from other ones?

What I mean is that every student has some abnormally high grades on their transcript, simply because some courses are extremely easy for one reason or another, or because some professors are overly easy graders. Is there any way for an employer to tell one set of inflated grades from another without assuming that because an applicant is black, that all of their grades are inflated because they are black? That strikes me, given the lack of evidence presented so far, as based solely on the prejudice that a black student couldn't have gotten those grades on their own.

As to your point about affirmative action...don't the argument you make and the example you cite deal with different things? To bring this back to our other discussion: let's assume for the sake of argument (since I'm not an expert on this issue) that blacks find it easier to get into the federal or state civil service. So what? Doesn't it matter more how they do while in the civil service, much as one's grades from a college matter more than the sheer fact of getting into a college? Also, do you have any evidence that other awards of the kind you bring up are "racially normed" to provide a rational the different value placed on those awards for whites and blacks?

To David Thomson: I'm sorry, that's just not going to cut it. First, you allege that such racially inflated grading is rampant at schools like Columbia and Harvard. Well, grade inflation is very present at schools like Columbia and Harvard (certainly I've gotten some inflated grades while a student at Columbia). So what? I made a specific claim about the nature of grading in colleges that would work against "racially norming" grades. Do you have any evidence or analysis to rebut that claim, other than talking about "hermits in the woods" or my school selling its soul to the devil? To say that Harvard is infamous for its grade inflation scandals means nothing unless you can prove that its grade inflation is only for certain groups (and I don't believe anybody is asserting that such is the case now).

Posted by: Mark on December 14, 2002 07:48 AM

To Tom Maguire: the problem with the idea that professors would grade certain groups of students differently based on race is that in any class larger than a seminar, the professor often does not know the race of his/her students. They see the faces every day, to be sure, but unless they actually have some sort of student-teacher relationship (which is virtually non-existent in lecture classes going by my experience thus far), they don't connect faces with names. It seems to me that your argument depends on this ability to match (black) faces with the names on a paper or an exam, and I don't see how that could happen in the lecture classes which make up the majority of college classes.

This is a wonderful example of the danger of threads getting away from their original subject! Because of course, according to the study which is the cause of all this discussion, there might be a way that professors can subconsciously or not guess who is black: by the names of the students. Another study is called for! Submission of the same papers with different names to unwitting professors. Let's see how the grad students mark them.

But I think this is balderdash in any case. Those who talk about blacks not "deserving" their place in higher education, do not seem to think anything about the guys and girls from Greenwich who pay big bucks for inside advisors (i.e. former admissions officers) to help them get in. Or that a poor student working 20-30 hours a week in the canteen, has less time and energy to study and fulfill requirements than the student supported by Mommy and Daddy with a big credit card.

Posted by: Andrew Boucher on December 14, 2002 08:37 AM

“Those who talk about blacks not "deserving" their place in higher education, do not seem to think anything about the guys and girls from Greenwich who pay big bucks for inside advisors (i.e. former admissions officers) to help them get in. Or that a poor student working 20-30 hours a week in the canteen, has less time and energy to study and fulfill requirements than the student supported by Mommy and Daddy with a big credit card. “

Life is not fair. You are, though, admittedly presenting an excellent argument why special classes should be offered to the borderline applicants. I always prefer employing sport analogies in these situations. Do you want to see someone on the team who is there only because of an affirmative action policies? Of course not, but why do so many say such patently foolish things when it comes to matters of the mind?

Posted by: David Thomson on December 14, 2002 09:46 AM

“Are you talking about admissions standards? That has _nothing_ to do with our discussion! We've established that affirmative action sets different standards for different groups. The discussion right now is about _grading_ not admissions. And if you're talking about exams, how would such a requirement exist? How would it be enforced? I made a specific claim based on my experience as a college student. Do you have any specifics to rebut that claim? “

I’m sorry but your attempting to distinguish between admissions standards and actual grades is utterly absurd. It is virtually impossible to do a study proving that professors don’t give more lenient grades to their black students. How would you even set up such a study? The unqualified students are now in the door and the corrupt leaders of Harvard and Columbia will make sure the dirt doesn’t come out. In some respects, the top officials of these schools are similar to the Catholic bishops. Any academic professional who tells the truth will likely destroy their career. I have every logical right to assume that the corruption didn’t end at the front door. It is your responsibility to prove otherwise.

“Well, grade inflation is very present at schools like Columbia and Harvard (certainly I've gotten some inflated grades while a student at Columbia). “

Thank you for not hiding the truth. This is most admirable. The prosecution rests its case.

Posted by: David Thomson on December 14, 2002 10:08 AM

Last but not least, we would never be having this debate if the Liberal radical did not drop the standards of their respective schools. The chickens are coming home to roost and they have nobody to blame but themselves. You should have never dropped the standards for minority candidates if you wanted the outside world to respect them. What were you thinking? At best, your well meaning sentiments messed up your ability to think and follow a logical argument.

Posted by: David Thomson on December 14, 2002 10:22 AM

Let's see, where to begin. Why is it my burden to prove that professors don't give students better grades than they deserve because of their race? I'm not demanding a scientific study or any such rigmarole; all I want is a specific mechanism by which this could happen. I've already given one reason as to why professors wouldn't grade by race (simply because they don't _KNOW_ the race of their students, because in any class larger than a seminar, they never have the opportunity to connect names and faces). All I want is for you to deal with that analysis.

How do professors know the race of their students? Why would they give minority students higher grades (or, given that the goal of affirmative action is at least in part to allow minorities increased _access_ to the institutions in question, why would professors have an incentive to give them higher grades?)? How on earth would the administration bring pressure to bear on an individual teacher who gave a black student a C? What is the incentive for a teacher to keep silent if such pressure occurred? Why is it my burden to defend against a counter-intuitive action, especially given that you've provided no mechanism for which this action would take place?

Oh, and one more thing. You say that the "prosecution" is able to rest its case because I've been in courses where I and others have benefited from grade inflation. That would work much better for the "prosecution" if I didn't happen to be white. Sorry.

Posted by: Mark on December 14, 2002 11:32 AM

I think the "reasoning" goes like this.

\begin{paraphrase}
1. Harvard and Columbia use racial preferences in admissions, accepting much weaker black students.
2. I am a believer in colorblind admissions, and I believe racial preferences (a/k/a affirmative action) is racist.
3. Grading students differently by race is racist.
4. "Therefore" Harvard and Columbia (might?) practice different grading by race.
\end{paraphrase}

On my planet, 1 and 3 are true, 2 I don't believe but honest persons might, and 4 is an outrage. Anyone making such a claim should back it up. I've never seen anything like it at four different universities.

Posted by: Andrew Lazarus on December 14, 2002 11:52 AM

Because I have no access to the figures, I have to rely on what I saw and heard at the time. Although it is not so now, it was then utterly commonplace for white professors to overgrade blackstudents. Any professor who did not overgrade blackstudents either felt the impulse to do so or saw others doing it. From that, I inferred a motive for overgrading white students, too.

HARVEY C. MANSFIELD, The Chronicle of Higher Education , April 6, 2001

Posted by: Bucky Dent on December 14, 2002 12:35 PM

I think the first clause of the second sentence is telling, given that Mansfield is talking about practices from a couple of decades ago.

Posted by: Mark on December 14, 2002 01:55 PM

I said that when grade inflation got started, in the late 60's and early 70's, white professors, imbibing the spirit of affirmative action, stopped giving low or average grades to black students and, to justify or conceal it, stopped giving those grades to white students as well.

...got started...

Ibid.

Posted by: Bucky Dent on December 14, 2002 02:35 PM

Mark 'n' Tom:
It probably varies from university to university, but at Columbia I doubt that the professors teaching large lecture courses do the grading. It's not true at the University of Pittsburgh, in the classes I've taught--the grading is done by teaching assistants, who teach the students in batches of 20 or so. On the other hand, sometimes the TA's divide up exams by problems rather than by student (so one TA will grade every student's question 1).

Anyway, since there is no explicit affirmative action policy in grading, why should we assume that all grading bias is in favor of black students? Perhaps some graders have an unconscious bias against black students--not that they're outright bigots, but that they subconsciously assume that black students benefited from affirmative action, and so look at their work with a more critical eye, and....

So I think Mark is right to say that the burden of proof is on those who want to prove that black students get more inflated grades than white students.

The Mansfield article isn't on point, by the way, because Mansfield is saying that affirmative action led to across-the-board grade inflation. Not to mention that his evidence is shaky.

(Tom M.--I couldn't figure out your argument until your fourth post or so--maybe that's why some thought you were fishing for reasons to object. Can't we all get along?)

Posted by: Matt Weiner on December 14, 2002 03:16 PM

As top schools overdraw their available pool of minority students, they
must dip into the students credentially suited for lower-tier colleges.

"The problems of mismatching and artificial failing proceed on down
the academic pecking order," Sowell writes. "Nationwide, 74 percent
of black students have failed to graduate, five years after entering college."

Sowell argues that most of those failures are artificial, caused by affirmative
action's mismatching the students with the schools.

When some schools recognize this problem, Sowell explains, they seek
for minority students. This move leads to the perception mentioned in the
Wall Street Journal and also causes highly qualified minority students to
defend their academic reputation needlessly to teachers and employers.

"High attrition rates of minority students on "white campuses"
is evidence of racism. [?]
Again, Sowell points to mismatching students with the
university's standards. At Berkeley, Sowell found that only 22 percent of Hispanic
period. He found similar figures for black students. "If the all-purpose
explanation is racism," Sowell asks, "then why did this racism have
such radically different effects on people of the same race with different test
scores?"

Posted by: Bucky Dent on December 14, 2002 03:57 PM

Tom, I'm not suggesting you're racist. I agree with Andrew Lazarus's comment.

Does the Sowell book contain any evidence that affirmative grading exists? He appears to cite David Riesman for it, and I can't find much of anything about him online.

Its supposed existence also has little or no relevance to the study under discussion. From page 18 of the study, only 11% of the ads had an education requirement (of ads that did have an education requirement, they broke down as 9% high scool, 49% some college, and 42% 4-year degree).

As they say on page 18, "In other words, these job requirements have no statistically discernible effect on the state of discrimination."

So, let's go over the given objections to the results of the study:

1) The study is detecting class bias, not race bias (appears to have been controlled for through zipcode & name usage based on SSN frequency).
2) The study didn't control for the effects of address (it did).
3) Preferential college admissions for minorities leads to employers discounting their degrees (irrelevant, only 11% of jobs even wanted education, and education requirements had no effect on the callback rate delta).
4) Preferential college grading for minorities leads to employers discounting their degrees (irrelevant, only 11% of jobs even wanted education, and education requirements had no effect on the callback rate delta).
5) Employers are adjusting for the perceived cost of discrimination claims (estimated cost is too low by a few orders of magnitude to explain a 50% difference in callback rates).
6) The tight labor market, or that the city voted for Al Gore, has something to do with it (relevance of these objections was not presented).
7) There's something subtle going on involving the african-american preference for new names vs. the caucasian preference for established names.

7 appears to be the only reasonable objection, but I can't figure out how an unfamiliar name would lead to a *50%* difference in callback rates, especially when you consider that no doubt a large component of the "unfamiliar name" mechanism is: surprise! racism.

Posted by: Jason McCullough on December 14, 2002 05:06 PM

5) Employers are adjusting for the perceived cost of discrimination claims (estimated cost is too low by a few orders of magnitude to explain a 50% difference in callback rates).

How may employees have you hired with your own capital at risk?

Posted by: Bucky Dent on December 14, 2002 06:16 PM

I liked this: "7) There's something subtle going on involving the african-american preference for new names vs. the caucasian preference for established names.

7 appears to be the only reasonable objection, but I can't figure out how an unfamiliar name would lead to a *50%* difference in callback rates, especially when you consider that no doubt a large component of the "unfamiliar name" mechanism is: surprise! racism."

Well, if you say it like this it doesn't sound so strange: "Interviews were requested for 10.1 percent of applicants with white-sounding names and 6.7 percent of those with black-sounding names, a total difference of only 3.4%."

If the authors had done a straight face-off between a Jethro Hatfield and a Lawrence Kennedy, would there have been a statistically significant difference in callback rates? If so, how much would it have been? In a nutshell, that's the issue.

Posted by: Anarchus on December 14, 2002 08:01 PM

It's disappointing to see that racism is still so rampant, and even more so to find that it pervades our very way of life in such a cloak-and-dagger means.

Posted by: RichardCross (Desi) on December 14, 2002 10:43 PM

Speaking of specious reasoning:

>>5) Employers are adjusting for the perceived cost of discrimination claims (estimated cost is too low by a few orders of magnitude to explain a 50% difference in callback rates).<<

1. The "estimated cost" is a joke. 799/800 claims are valid, says a government agency whose existence is predicated on its finding racial discrimination cases to litigate. What matters to a business are actual costs, which they (at least larger businesses) would undoubtedly have handy from their accounting departments.

2. The rejection rates are 90% for neutral names, and 93% for "black-sounding" names. Which, if we can believe this study, could be brought back to parity by the latter using initials, rather than names. This hardly seems something to worry about.

Further, the authors have no way of knowing how many, what quality, or from whom, other resumes there were for their make-believe resumes to compete with.

They also do not know what the racial make-up of the employers' personnel depts. were. Nor the racial composition of the employers' current workforce (and hence the racial targets that employers subject to Affirmative Action "goals", "time tables", or "quotas", just might be legally compelled to try to meet).

After all, "Racial Discrimination" is the official policy of almost all government in America.

As for:

>> 6) The tight labor market, or that the city voted for Al Gore, has something to do with it (relevance of these objections was not presented).<<

It wasn't presented because it's obvious that the labor market immediately post 9-11-01 might not be telling us anything worth knowing. Does Jason suspect that all the airlines in Chicago would have been doing a lot of hiring in that period? Or that the competition to find productive employees would have been as vigorous as, say, from 1995-2000?

As for Al Gore, his approving view of racial discrimination in hiring is not shared by your typical American. Presumably, in cities that voted for him there might be a disproportionate number of people with atypical views of the role race should play in hiring. Meaning that racial targetting might be more prevalent in those cities.

Posted by: Patrick R. Sullivan on December 15, 2002 09:26 AM

'The "estimated cost" is a joke.'

Provide your own cost estimate, then. I want to see this devastating incidence of discrimination claims.

'The rejection rates are 90% for neutral names, and 93% for "black-sounding" names. Which, if we can believe this study, could be brought back to parity by the latter using initials, rather than names. This hardly seems something to worry about.'

So blacks should just hide their race, and the racism disappears. Great.

'It wasn't presented because it's obvious that the labor market immediately post 9-11-01 might not be telling us anything worth knowing. Does Jason suspect that all the airlines in Chicago would have been doing a lot of hiring in that period? Or that the competition to find productive employees would have been as vigorous as, say, from 1995-2000?'

What relevance would this have to the racial distribution of callbacks?

'As for Al Gore, his approving view of racial discrimination in hiring is not shared by your typical American. Presumably, in cities that voted for him there might be a disproportionate number of people with atypical views of the role race should play in hiring. Meaning that racial targetting might be more prevalent in those cities.'

Which means that in other cities, the results should be even worse. Right?

Posted by: Jason McCullough on December 15, 2002 03:03 PM

Provide your own cost estimate, then.

Regulatory costs are always impossible to calculate with precision. That inability does not magically render those costs, or their unintended side effects, moot.

I'll repeat my question: How may employees have you hired with your own capital at risk?

Posted by: Bucky Dent on December 15, 2002 04:21 PM

"I'll repeat my question: How may employees have you hired with your own capital at risk?"

This is an especially important question when you realize that most business owners probably are lucky to earn around $60,000 annually (does anyone have the specific average?). They do not possess the deep pockets of a major corporation. Therefore, even a minor legal flap can be very costly. A discrimination claim that doesn’t get past first base might still cost a couple hundred dollars. Guess what, that’s a painful punch to the solar plexus for someone who’s may be lucky to net$700 a week. My figures admittedly are rough guesses, but I suspect that I’m in the proverbial ballpark.

Posted by: David Thomson on December 15, 2002 05:10 PM

'Regulatory costs are always impossible to calculate with precision. That inability does not magically render those costs, or their unintended side effects, moot.'

So, the costs are big, but we don't know what they are. But they're big, and uh......

Posted by: Jason McCullough on December 16, 2002 12:19 PM

IOW, the number of people you have hired with your own capital at risk - a proxy for direct knowledge of what you're talking about - is equal to the number of Bell Curve chapters you've read - a book you've condemned without reading.

Posted by: Bucky Dent on December 16, 2002 12:45 PM

I don't quite see what Mr. McCullough's experience or lack thereof as a businessman has to do with the question he's asking. You've made a simple contention, that the costs of discrimination lawsuits are so high that employers are (justifiably) leery of hiring someone who might file such a suit. He's responded by asking you to cite specific numbers or at the very least ballpark figures to back up your point. You've answered that you can't really calculate those costs with any degree of accuracy, but that they exist and their effects are real. So if you can't calculate them with any kind of precision, why should we believe your assertion? David Thomson has at least started to provide some kind of basis for discussion w/his latest post (though I would prefer actual economic statistics rather than "rough ballpark estimates." Can you do the same?

Posted by: Mark on December 16, 2002 12:51 PM

"You've made a simple contention, that the costs of discrimination lawsuits are so high that employers are (justifiably) leery of hiring someone who might file such a suit."

I actually took the argument a step further. My premise is that the cost of these regulations is a relative term. The answer might be totally different if we are discussing a Fortune 500 company, or a mom and pop business. Reality check: most businesses in the United States are of the latter kind! A so-called “minor” legal complaint can be a hair raising experience for a business owner barely grossing \$60,000 annually. The mere fact that you might have to contact your attorney may destroy that week’s earnings. Responding to a discrimination charge will probably cost you a minimum of two hours of your precious time.

Posted by: David Thomson on December 16, 2002 01:15 PM

The self-evident "chilling effect" of litigation fear can't be captured in aggregate data. I think that's obvious.

Equally obvious is the perspective and self-interest of the person whose capital and reputation are at risk in the hiring process.

One could argue that the effect is captured by the callback data cited, and that it is being perjorativeively misidentified as "racial prejudice".

Posted by: Bucky Dent on December 16, 2002 01:20 PM

Jason's cost calculation was utterly unbelievable. Here's another problem: if the discrimination claim reaches the EEOC, you can be sure that the company and the plaintiff have discussed it extensively and have not solved the problem informally. Therefore, it's probably legitimate to assume that the company doesn't agree with the claim. The real measure of how destructive the threat of litigation can be (let's leave aside the real benefits of the process) is how many claims of discrimination are made for each minority employee. The claims don't have to reach litigation, either. This is going to be very hard to quantify but it's reasonable to assume it is a significant concern.

Posted by: JT on December 16, 2002 02:25 PM

"And, "Tamika" (94.6% non callback rate, 8th worst of 26 names), I would associate with Japanese, not African-American."

"How do you think Ardvard Ramadan would do. I am extremely liberal but Aisha what kind of name is that. Sounds Indian to me."

I'd make a joke on the order of "I think you're confusing her for Hind al-Hunud," but... What could possibly be more traditional than *Aisha*? Ishtar? Enheduanna?

And Japanese? Honestly, I cannot tell if this is disingenuousness or evidence that the US is far more segregated than I ever imagined... not to mention the gentleman has a sorry grasp on the oeuvre of DMX.

But what I really want to know is, if employers are reluctant to employ African-Americans because they might file a lawsuit, are they petrified of hiring straight white men because this is a group most likely to cause the firm huge losses by making inappropriate comments about "jellybeans" and/or groping a secretary?

I am extremely liberal myself, so I will assume that this is some sort of Cointelpro-esque disinformation campaign to impugn Mr.deLong's site.

Posted by: Drapetomaniac on December 16, 2002 03:06 PM

'IOW, the number of people you have hired with your own capital at risk - a proxy for direct knowledge of what you're talking about - is equal to the number of Bell Curve chapters you've read - a book you've condemned without reading.'

You're entirely correct: they both don't have a damn thing to do with the discussion at hand. Maybe math has previously unknown parallels to the weight military service carries in defense policy debates? ("As someone who's practiced statistics over in the Vietcong Delta, Senator.....")

'But what I really want to know is, if employers are reluctant to employ African-Americans because they might file a lawsuit, are they petrified of hiring straight white men because this is a group most likely to cause the firm huge losses by making inappropriate comments about "jellybeans" and/or groping a secretary?'

Obviously more research is called for!

Anyway, sure, there's reasonable arguments that potential discrimination claims have costs beyond what's directly visible, and could lead to noticable gaps in hiring. However:

1) I've never seen an attempt to quantify these invisible costs.
2) It's a rather annoying debating tactic to simultaneously insist that these costs can't be quantified, but you can quantify them well enough to decide that both my attempt is "silly", and you're quite certain the costs explain a significant portion of a 50% gap in callback rates to boot.
3) The costs would need to be awfully big to explain a mean job search for identically qualified african-americans being half again as long as that for whites (a reasonable extrapolation from the results of the study, I think).

'My premise is that the cost of these regulations is a relative term. The answer might be totally different if we are discussing a Fortune 500 company, or a mom and pop business.'

Regression to the mean should drive the incidence of claims at a large company towards the average claim rate, yes; they'd have a lower risk premium than small companies, due to the smaller beta. (is that the right term?) That's the only difference that I can come up with, though, pointing towards aversion to discrimination claims being sublinear to increasing company size.

An enterprising researcher out there *should* be able to construct a risk pricing model for these claims at various types of companies, present a reasonable estimate of the increased costs per worker, and end up with something resembling a "discrimination aversion" number. Google doesn't give much, though. Anyone?

Posted by: Jason McCullough on December 16, 2002 04:15 PM

"Civil rights laws were not passed to protect the rights of white men and do not apply to them."
-- Mary Frances Berry, Chairwoman, US Commission on Civil Rights

Unbelievable! (Rhetorically speaking...not literally.) Trent Lott is--properly, in my opinion--likely to lose his Senate Majority Leader status for saying less (in my opinion).

So much for "equal protection of the law." (Interesting...I guess the 14th amendment does apply to the states...not the federal government!)

Posted by: Mark Bahner on December 16, 2002 04:32 PM

In a country where AOL, Merrill Lynch, American Express and the State Department are led by black men, and the National Security Council is led by a black woman, the notion that hiring considerations are made in accordance to "racial prejudice" is not an empircally serious conclusion.

With widespread, high-profile black success, employers, who always look for an economic edge, must have an *economic* reason for hesitiating to call back LaTeesha instead of of Lynn.

Assuming they're simply wicked is the least likely answer.

Posted by: Bucky Dent on December 16, 2002 05:08 PM

3) The costs would need to be awfully big to explain a mean job search for identically qualified african-americans being half again as long as that for whites (a reasonable extrapolation from the results of the study, I think).

No. The RISK of a single high-cost event can explain the phenomenon at the staffing desk level.

And the excruciatingly clear parallel to the unread Bell Curve is this: You can't take seriously the arguments of someone who literally has no understanding of the basic issues under discussion.

Condemn an unread book? Smear people who do a job you can't possibly understand for a "thought crime" you infer from fragmented and incomplete knowledge?

Those are fine tactics for tabloid editorialists, but poor ones for the real world.

Posted by: Bucky Dent on December 16, 2002 05:21 PM

Some thoughts on a heated thread. I confess that I came to this late and glossed over some of the later posts. If in fact others have made these points, please excuse me.

The first point is made quite clearly by Tom Kane in e.g., the Black-White Test Score Gap volume cited above: One cannot infer from blacks' lower average SAT scores at Harvard alone that they were admitted under a lower standard. It may reflect in part differences in the distribution of scores above whatever threshold you use. For example, if Harvard admitted anybody with an SAT above 1000--regardless of race--African-Americans would still have lower average SATs. This is simply because, even above 1000, blacks are disproportionately represented among "low" scores (e.g., 1000-1200).

Second point: In the discussion of whether blacks are more likely than whites to file discrimination claims, this unconditional question seems irrelevant. Nonetheless, many here seem to focus on it.

This question seems to ignore the fact that minorities and women might be more likely to face actual discrimination than whites or men. Indeed, couldn't the fact that these suits are being filed might suggest that, at least in 1/3 of the cases, there is evidence of discrimination.

To draw a silly example, one cannot infer that just because a butcher is more likely than a stockbroker to file a workers comp claim for a severe hand injury that butchers are more likely file false workers compensation claims for such an injury.

The question an employer might ask is whether blacks are more likely to file a suit conditional on the employer discriminating against the employee and similarly whether they are more likely than whites to file unmerited lawsuits, in which the employer does not discriminate. From the numbers cited above we know only that 2/3 of all suits are tossed out. What seems relevant to me is whether this differ by race or sex, for example?

Posted by: Brian on December 16, 2002 05:57 PM

I guess it has to be reduced to kindergarten level, keep in mind:

>> "Civil rights laws were not passed to protect the rights of white men and do not apply to them."
-- Mary Frances Berry, Chairwoman, US Commission on Civil Rights <<

Now, put yourself in the place of a potential employer whose OWN MONEY is riding on his decision to hire one of two people with exactly the same qualifications: (1)A white male whose potential danger to you as a civil rights litigant is exactly ZERO, according the the expert Ms Berry. Or, (2) A black female whose potential danger is greater than zero.

Who is the less dangerous hire?

BTW, has Jason ever heard of a "representative sample"?

Posted by: Patrick R. Sullivan on December 17, 2002 07:08 AM

>>But what I really want to know is, if employers are reluctant to employ African-Americans because they might file a lawsuit, are they petrified of hiring straight white men because this is a group most likely to cause the firm huge losses by making inappropriate comments about "jellybeans" and/or groping a secretary? <<

How do you know white males are more likely to do either?

Posted by: Patrick R. Sullivan on December 17, 2002 07:12 AM

"Civil rights laws were not passed to protect the rights of white men and do not apply to them."
-- Mary Frances Berry, Chairwoman, US Commission on Civil Rights"

I am so glad that you brought up the subject of the vile Mary Frances Berry. This woman is a disgusting hate monger who should be disowned by the Democrat Party. When are we going to hear calls for her ouster? Berry may very well be the most obnoxious individual in Washington politics. Democrats should be deeply ashamed that she is a member of their party.

Posted by: David Thomson on December 17, 2002 07:21 AM

>> And Japanese? Honestly, I cannot tell if this is disingenuousness or evidence that the US is far more segregated than I ever imagined <<

http://www.20000-names.com/female_japanese_names.htm

>> TAMIKA: people <<

Posted by: Patrick R. Sullivan on December 17, 2002 07:25 AM

Jason writes: "It's a rather annoying debating tactic to simultaneously insist that these costs can't be quantified, but you can quantify them well enough to decide that both my attempt is "silly", and you're quite certain the costs explain a significant portion of a 50% gap in callback rates to boot."

I sympathize with Jason's complaint here but he is distorting the argument. The argument is that some or all of the disparity in calling back applicants may be explained by the perception (key word there -- pay attention) that certain ethnic categories are more likely to file discrimination lawsuits. In this case, racism wouldn't be driving the lower rate of callbacks; it would be a rational response to a business issue.

Jason: for just a second, drop your ideological position, which I believe is driving your argument, and consider that a mixture of litigiousness and civil rights law might be actively detracting from an ethnic group's economic advancement. It's possible isn't it?

Brian: thanks for carefully restating the position. I don't think you've added anything to the argument, though. The assumption is that the employers are not avowed racists and believe they follow US law and don't discriminate according to race.

Posted by: JT on December 17, 2002 09:03 AM

JT: Whether employers discriminate by race or not is one of the central questions of the Bertrand/Mullainathan paper, though. In my opinion, even with its flaws, this and other papers put the burden of proof on those who claim that there is no discrimination. Either way, assuming that employers do not discriminate makes much of the debate here uninteresting, doesn't it?

Again, I am not sure how--once you allow the possibility that some employers discriminate--you can separately identify the degree to which minorities are more likely to file frivolous suits versus the extent to which they are more likely to face actual discrimination. I would have to give it more thought and welcome yours.

Posted by: Brian on December 17, 2002 10:28 AM

I am not sure how--once you allow the possibility that some employers discriminate--you can separately identify the degree to which minorities are more likely to file frivolous suits versus the extent to which they are more likely to face actual discrimination.

It is exactly this grey zone that acts like an ideological Rorschach test: The market types say economic incentives drive any reported under-hiring of blacks, while social liberals cite disparities as proof of racist bigotry.

My personal and unprovable read is that bigotry per se is on the decline (see Powell, Rice, Oprah, etc.), but not at zero. Unfortunately, well-intentioned efforts to right wrongs have introduced factors making it riskier for some enterprises to hire minorities, making "discriminiation" economically rational where it otherwise would not be.

Short of a thought police survey, the relative weights of these factors cannot be exactly determined. But it would be wrong to say either doesn't exist.

Posted by: Bucky Dent on December 17, 2002 10:43 AM

"In my opinion, even with its flaws, this and other papers put the burden of proof on those who claim that there is no discrimination."

I agree.

"Either way, assuming that employers do not discriminate makes much of the debate here uninteresting, doesn't it?"

I think Bucky makes the point very well: no one's assuming there's no discrimination. If my argument was an attack on anything, it was aimed at the legal infrastructure that exists to remedy discrimination complaints.

"Again, I am not sure how--once you allow the possibility that some employers discriminate--you can separately identify the degree to which minorities are more likely to file frivolous suits versus the extent to which they are more likely to face actual discrimination. I would have to give it more thought and welcome yours."

Good question. It's not clear that there's an objective solution.

Posted by: JT on December 17, 2002 12:51 PM

"Civil rights laws were not passed to protect the rights of white men and do not apply to them."

An interesting fact: Google does not produce a single link documenting where/how/why this statement was made; it's a blind quote on every single match. Does anyone have better background on this?

'Jason: for just a second, drop your ideological position, which I believe is driving your argument, and consider that a mixture of litigiousness and civil rights law might be actively detracting from an ethnic group's economic advancement. It's possible isn't it?'

I'm sure it exists, but as I pointed out, I don't see how it's a significant factor. Unless someone produces a better estimate, I stand by mine.

'Now, put yourself in the place of a potential employer whose OWN MONEY is riding on his decision to hire one of two people with exactly the same qualifications: (1)A white male whose potential danger to you as a civil rights litigant is exactly ZERO, according the the expert Ms Berry. Or, (2) A black female whose potential danger is greater than zero.'

If you have two identical candidates in front of the same recuiter, excepting race, and they can only call one, yes, they will rationally always call the one least likely to file discrimination claims. I'm not arguing that point.

What I am arguing is that the productivity adjustment for minority employees due to their increased employment cost (the end result of these claims) is nowhere near large enough to explain the seen results.

What kind of productivity gap is necessary to result in a 50% longer search time? Is there a model out there for matching the ratio of actual/required productivity to mean time to hire?

Posted by: Jason McCullough on December 17, 2002 04:28 PM

"An interesting fact: Google does not produce a single link documenting where/how/why this statement was made; it's a blind quote on every single match. Does anyone have better background on this?"

"MARY FRANCES BERRY (current head of U.S. Commission on Civil Rights)—"Civil rights laws were not passed to protect the rights of white men and do not apply to them." [Civil Rights Under Reagan, San Francisco, ICS Press, 1991, p. 141.]"

http://216.247.220.66/archives/racerelations/lubinskas11-24-99.htm

Posted by: David Thomson on December 17, 2002 04:53 PM

Gosh, there's even more where that came from:

NOTE: On Rivera Live Monday night, black talk-show host Tavis Smiley denied that there is anti-white hatred or prejudice in America. Some counter-examples:

SONNY CARSON (black activist in New York, when asked if he was anti-Semitic) "I am anti-white. I don't limit my 'anti' to just one group of people." [Mark Mooney, "Ex-Dinkins Organizer Boasts He's 'AntiWhite'" New York Post, October 21, 1989, p. 3.]

MILES DAVIS (black jazz musician) "If somebody told me I had only one hour to live, I'd spend it choking a white man. I'd do it nice and slow." ["Miles Davis Can't Shake Boyhood Racial Abuse," Jet, March 25, 1985.]

ELDRIDGE CLEAVER (former Black Panther leader on why he raped white women) "Rape was an insurrectionary act. It delighted me that I was defying and trampling upon the white man's law, upon his system of values, and that I was defiling his women." [Eldridge Cleaver, Soul on Ice, McGraw-Hill, 1968, p.14.]

GUS SAVAGE (former U.S. Representative from Chicago to a white member of the press) "I don't talk to you white motherf— .… You bitch motherf— in the white press…. F— you, you motherf—ing assh— … white devils." [Marilyn Rauber, "Reporter Says Black Rep Hurled Racial Slurs," New York Post, June 27, 1991, p. 18.]

CHINO WILSON (in an editorial in the Daily Collegian, campus newspaper at Penn State University) "After looking at all the evidence there is only one conclusion: white people are devils…. I believe that we must secure our freedom and independence from these devils by any means necessary, including violence…. To protect ourselves we should bear arms (three handguns and two rifles, maybe an M-16) immediately and form a militia. …. So black people, let us unite, organize, and execute." [Chino Wilson, "African American Students Should Not Trust 'Devilish' White People," The Daily Collegian, Penn State University, January 28, 1992.]

KHALID ABDUL MUHAMMED (former assistant to Louis Farrakhan—current leader of the New Black Panther Party) "Hollywood is owned by these so-called Jews. Look at the movies they make about us, black people killing black people. Let's make some revolutionary movies where we kill white people in the movie. Kill 'em so hard you have to cover up your popcorn from the blood spraying out of the screen." [Speech at San Francisco State University, May 21, 1997.]

KHALID ABDUL MUHAMMED (on what South African blacks should do to any whites who refuse to leave South Africa): "We kill the women. We kill the babies. We kill the blind. We kill the cripples. We kill them all…. When you get through killing them all, go to the goddamn graveyard and kill them a-goddamn-gain because they didn't die hard enough."[November 29, 1993 speech at Kean College in Union, New Jersey.]

MARY FRANCES BERRY (current head of U.S. Commission on Civil Rights)—"Civil rights laws were not passed to protect the rights of white men and do not apply to them." [Civil Rights Under Reagan, San Francisco, ICS Press, 1991, p. 141.]

AUGUSTIN CEBADA (Head of the Brown Berets, a Hispanic activist organization at a July 4, 1996 rally)—"We're here today to show L.A., show the minority people here, the Anglo-Saxons, that we are here, the majority, we're here to stay. We do the work in this city, we take care of the spoiled brat children…. we are the majority here and we are not going to be pushed around."

AUGUSTIN CEBADA "Go back to Simi Valley, you skunks! Go back to Woodland Hills! Go back to Boston! Go back to Plymouth Rock, Pilgrims! Get out! We are the future. You are old and tired. Go on. We have beaten you. Leave like beaten rats. You old white people, it is your duty to die." [Quoted in Barbara Coe, Reconquista, The Takeover of America, California Coalition for Immigration Reform, 1998, p. 20.]

PROF. JOSE ANGEL GUTIERREZ (University of Texas, Arlington) "We have an aging white America. They are dying. They are sh—ing in their pants with fear!… I love it!"—[Speech of Jan. 1995, quoted in Coe, Reconquista, p. 16.]

SUPREME COURT JUSTICE THURGOOD MARSHALL (in a conversation with Justice William Douglas about racial preferences) "You guys have been practicing discrimination for years. Now it is our turn." [William O. Douglas, The Court Years 1939-1975, New York, Random House, 1980.]

BELL HOOKS (black professor of English at City College of New York) "I am writing this essay sitting beside an anonymous white male that I long to murder." [From her book A Killing Rage, quoted by David Horowitz in Hating Whitey, Spence Publishing, 1999, p. 31.]

SISTER SOULJAH (rap artist and black activist) "If black people kill black people every day, why not have a week and kill white people." [R.W. Apple "Jackson Sees 'Character Flaw' in Clinton's Remarks on Racism, New York Times, June 19, 1992.]

ICE CUBE (black rapper and actor, on the anti-Korean album Death Certificate) "So don't follow me up and down your market. or your little chop suey ass will be a target. So pay your respects to the black fist or we'll burn your store right down to a crisp." [Eric Briendel, "Rap Star to Koreans: 'We'll Burn Your Stores,' " New York Post, Dec. 5, 1991, p. 29.]

AMIRI BARAKA (black poet and writer) "You cant steal nothin from a white man, he's already stole it he owes you anything you want, even his life. All the stores will open up if you will say the magic words. The magic words are: Up against the wall motherf— this is a stick up!" [Quoted in Anne Wortham, The Other Side of Racism, Ohio State University Press, 1981, p. 257.]

MARIO OBLEDO—(1998 Presidential Medal of Freedom recipient and former head of Mexican American Legal Defense and Education Fund—MALDEF) "California is going to be a Mexican state, we are going to control all the institutions. If people don't like it they should leave." [Tom Leykis Radio Show, June 7, 1998.]

MALCOLM X—"The death of over 120 white people is a very beautiful thing." [Speech in Los Angeles on June 3, 1962 upon learning of a plane crash. He also said on numerous occasions, "The white man is the devil."]

REV. JAMES CONE—"What we need is the destruction of whiteness, which is the source of human misery in the world." [Quoted in David Horowitz, Hating Whitey, Spence Publishing, 1999, p. 44.]

ART TORRES (former chairman, California Democratic Party)—"Remember, [Proposition] 187 [the measure to cut public benefits to illegal aliens] is the last gasp of white America." [The Social Contact, Summer 1998, p. 290.]

WILLIE BROWN (Mayor of San Francisco, to a white parent complaining that affirmative action would penalize his children) "I don't care about your idiot children." [The Social Contract, Summer 1998, p. 290.]

E N D

Posted by: David Thomson on December 17, 2002 04:58 PM

Memo to Jason: After learning to criticize books *after* you've examined them, and allowed for solutions other than thought crimes to explain social pheonomena, study The First Rule of Holes.

Posted by: Bucky Dent on December 17, 2002 05:09 PM

>> What I am arguing is that the productivity adjustment for minority employees due to their increased employment cost (the end result of these claims) is nowhere near large enough to explain the seen results.

>> What kind of productivity gap is necessary to result in a 50% longer search time? Is there a model out there for matching the ratio of actual/required productivity to mean time to hire?<<

Jason, the gap is only 3%, as I have continually pointed out to you. Neutral names (we don't have to think very hard to come up with blacks named Michael, Charles, Ken, Barry....) are rejected 90% of the time. As opposed to 93% for "black-sounding names", some of which have Islamic connotations that can't have been helpful after 9-11 (just as Gunter, Gerhard, and Dieter weren't probably helped by the sinking of the Lusitania).

And your estimate of the costs is, as I said, a joke. The costs are real to the employer whether someone considers the litigation "frivolous" or not. You are aware that the purpose of a business is to make money?

And the authors of this study are overselling their results when they write:

"Our evidence suggests that discrimination is an important factor in why African Americans do poorly in the labor market."

Their evidence shows no such thing.

Posted by: Patrick R. Sullivan on December 18, 2002 08:13 AM

Yes, I found the citation, but for something so shocking you'd think they would document what case/interview/general context she's talking about. I'm just a tad bit suspicious of quotes which only show up on "scary black quote" pages, without any further information.

Jason, the gap is only 3%, as I have continually pointed out to you. Neutral names (we don't have to think very hard to come up with blacks named Michael, Charles, Ken, Barry....) are rejected 90% of the time. As opposed to 93% for "black-sounding names", some of which have Islamic connotations that can't have been helpful after 9-11 (just as Gunter, Gerhard, and Dieter weren't probably helped by the sinking of the Lusitania).

Is it coincidental you frame the results as "90% vs. 93%, " Patrick? I'm not sure where you got that neutral-name callback rate, as I don't see it in the study, but surely you know that the given numbers in the study (resumes with white names have a callback rate of 10.08%, the same resumes with black names have a callback rate of 6.7%) mean that the chance of getting a callback for a given resume is 1 in 10 vs. 1 in 15.

Using these chances (and the magic of rounding for readability), it takes a white applicant 7 resume submissions, and a black applicant 10, to have a 50% chance of getting one or more callbacks. A 75% chance takes 13 resumes for a white applicant and 20 resumes for a black applicant. A 90% chance of one or more callbacks takes 22 vs. 33. In other words, that "3% gap" of yours is a bit of statistical obfuscation; a 50% gap in the relative chance of getting a callback (6.7% vs. 10%) results in: surprise! a roughly 50% gap in the number of resumes that need to be sent to get the same level of employer response.

And your estimate of the costs is, as I said, a joke. The costs are real to the employer whether someone considers the litigation "frivolous" or not. You are aware that the purpose of a business is to make money?

In other words, the costs for employers in aggregate are what you, a singular employer, think they are. Thanks for your contribution to the discussion.

Posted by: Jason McCullough on December 18, 2002 03:05 PM

the costs for employers in aggregate are what you, a singular employer, think they are

Exactly EXACTLY right!

Each employer makes their own decision based on their own cost/benefit risk/reward analysis. The aggregate data, insofar as it is accurate, simply sums that process up.

Posted by: Bucky Dent on December 18, 2002 04:15 PM

>> Is it coincidental you frame the results as "90% vs. 93%, " Patrick? I'm not sure where you got that neutral-name callback rate, as I don't see it in the study...<<

I got the numbers by passing 4th grade arithmetic. The rejection rate is every bit as relevant as the call-back rate...unless you've a predisposition to a specific conclusion. Rejection is the norm.

>> In other words, the costs for employers in aggregate are what you, a singular employer, think they are. Thanks for your contribution to the discussion. <<

This is astoundingly obtuse. The only costs that matter to an employer are HIS COSTS. The costs to the aggregate of employers are not subtracted from HIS revenues. As I said, you do understand that the purpose of a business is to make money?

Posted by: Patrick R. Sullivan on December 18, 2002 05:08 PM

That's it, I give up. Continue living in your little spin-doctored fantasy land, where your personal cost of discrimination claims is equivalent to the mean cost, and a 50% gap in mean time to hire is tiny.....

Posted by: Jason McCullough on December 18, 2002 05:31 PM

Speaking of living in fantasyland, people who bear the costs of their decisions usually don't waste time with nincompoop theorizing. There's a big wide world out there, Jason, try learning something about it.

Posted by: Patrick R. Sullivan on December 19, 2002 09:46 AM

I think racism is alive and well.

The solution is to work hard.

Look at the Chinese, at one time treated badly.

Now most of our products say "MADE IN CHINA".

Hard work = SUCCESS.

Posted by: Tom Paris on January 14, 2003 06:48 PM

There are always those predisposed to believing that racism never occurs, while others always see it. I think it was a Sociology professor in New York in the 70s that asked (to white people) something like:

***If you magically were to turn black. How much money would you pay to become white again?*****

One rational way to answer is to calculate the present value of the estimated lifetime loss in income due to discrimination. For those that believe there is no discrimination this approach would lead to zero. How would others look at this?

Posted by: Mike Mays on January 15, 2003 12:06 PM

There are always those predisposed to believing that racism never occurs, while others always see it. I think it was a Sociology professor in New York in the 70s that asked (to white people) something like:

***If you magically were to turn black. How much money would you pay to become white again?*****

One rational way to answer is to calculate the present value of the estimated lifetime loss in income due to discrimination. For those that believe there is no discrimination this approach would lead to zero. How would others look at this?

Posted by: Mike Mays on January 15, 2003 12:06 PM

There are always those predisposed to believing that racism never occurs, while others always see it. I think it was a Sociology professor in New York in the 70s that asked (to white people) something like:

***If you magically were to turn black. How much money would you pay to become white again?*****

One rational way to answer is to calculate the present value of the estimated lifetime loss in income due to discrimination. For those that believe there is no discrimination this approach would lead to zero. How would others look at this?

Posted by: Mike Mays on January 15, 2003 12:06 PM

A similar study showed that a lady was perceived as prettier if she had a name like Denise, or Heather, and perceived as less attractive if she had a name like Margot or Ursula.

This study isn't perfect, but it's interesting.

The question we should be asking is "how could a better study be designed?"

Posted by: Greg on January 15, 2003 07:34 PM