Tuesday, March 21, 2006

Help Wanted II - The Wait

I submitted the opinion piece to Detroit Free Press this morning. Perfect timing, as it turns out. Just yesterday, the Michigan Supreme Court declined to hear the Angry Left's argument against the referendum signatures.

Thanks again for all your help -- it came out far tighter than it went in.

Given the paper's overt bias against the referendum, though, I doubt it will see the light of MSM day.


At the moment, despite the Angry Left's best efforts, Michigan will have a referendum on the ballot this coming November to end affirmative action in state hiring and college admissions.

Predictably, the chattering classes have taken a firm stand against such a thing, no matter that they are also taking a stand against the US Constitution. Consequently, the local paper has been inundated with Op Eds opposing the referendum, with scarcely a thing to balance the argument.

This is where I need your help some more -- thanks to some excellent suggestions, and coaching from our resident professional journalist, I have hacked away at the original.

So, with submission to the Detroit Free Press pending, lock and load.


The Principled Case Against Affirmative Action

Judging from op-ed comments from the Detroit Free Press and Governor Granholm, affirmative action is fairness incarnate, and those who disagree carry something between a whiff and outright stench of racism.

Nothing could be further from the truth. The arguments against affirmative action are based on fundamental notions of fairness, dealing with problems rather than symptoms, avoiding institutionalized racism, and adherence to the Constitution.

In order to gauge the lie of the land, it is important to keep in mind why affirmative action exists in realms such as college admissions and public sector hiring.

The reason is singular and simple: on average, African-Americans' academic and job qualifications significantly trail those of other groups. If the person doing the selecting had nothing else to go on, they would be profoundly under represented in elite universities and public sector employment.

So what has affirmative action's opponents so exercised? How could they be against something so obviously good as diversity without being racists? In fact, the argument against group preferences is both principled and multifaceted.

Affirmative action deals with symptoms, not problems. Taking university admissions as an example, the problem is inadequate academic preparation, not racist admissions boards. By treating the symptom -- acceptance rate -- affirmative action reduces the pressure on fixing the real problem: frequently awful schools that afflict so many African-American children. At the individual level, affirmative action reduces the incentive to gaining the very academic credentials that are at the heart of the problem. Using a metaphor particularly apt for Detroiters, this is like using Bondo to deal with rust. Plastering over the symptoms does nothing to fix the rot below.

Equally galling, group preferences of any kind constitute a governmental spoils system akin to a lottery, where a few rewards are scattered arbitrarily among a larger number of applicants, all of whom are equally "eligible," since the only requirement is the correct skin color. For those who have trouble with irony, this is an excellent example. By requiring governmental racial classification, this very emphasis on skin color causes affirmative action to perpetuate the very racism against which it is ostensibly a bulwark.

What is more, since affirmative action defines diversity racially, it produces results indefensible to anyone not racially motivated. Clearly, the students who manage to graduate from Detroit's schools faced profound obstacles that are a legacy of the institutionalized racism once directed at African Americans. Equally, students who manage to graduate from schools in West Virginia's coal mining country also faced profound obstacles. Affirmative action's proponents, for whom diversity is solely a matter of skin color, would favor the former over the latter. In what way is that fair?

So far, the arguments against affirmative action have focused at the "macro" level. What are the effects on its beneficiaries? Since affirmative action at elite schools lowers the academic bar, unless admission requirements are wholly arbitrary, one might expect those admitted under its auspices to have lower graduation rates. California illustrates the case nicely.

After voters there prohibited group preferences, African-American college enrollment, overall unchanged, shifted sharply away from the prestige universities, just as affirmative action's proponents feared. However, this shift also affected graduation rates. Far more blacks were succeeding at schools like Cal-State LA rather than dropping out of, say, Stanford.

By getting students into schools for which they are academically unprepared, group preferences provide as good an example of unintended consequences as one is likely to find.

Finally, there is the little matter of the Constitution, the gold standard against which to judge affirmative action. The Ninth Amendment’s language and intent could not possibly be more clear: "No state shall make or enforce any law which shall … deny to any person within its jurisdiction the equal protection of the laws." Michigan can either uphold the Constitution, or enforce affirmative action. It cannot do both.

America's ideals lead, through the imperfect means of human agency, to the kind of society-wide fairness that comes from equal application of the law. Again with respect to college admissions, whether automatically admitting some top percentage of high school graduates, or providing admissions advantage to everyone from economically challenged zip codes, there are non-racial ways of obtaining student bodies with diverse backgrounds. Conversely, group preferences -- a euphemism for racism -- were immoral when directed against African-Americans; they are no less immoral when pointed the other way. It is time to cast group preferences once and for all into the deep blue sea and deal with problems, rather than focusing merely on symptoms.


Blogger Harry Eagar said...

'lie of the land' not 'lay of the land'

You're singing my song, here. Maybe tomorrow I'll have time to go over it carefully, particularly with a view to making cuts. It's long for a newspaper op-ed (unless the Detroit papers are more generous with space than I remember).

March 21, 2006 7:32 PM  
Blogger Bret said...

Unfortunately, this is not really a topic I can pretend to know much about. Also, I'm not sure what the proponents exact arguments are other than "They say Jim Crow. We say, hell no" - it's a little tough to argue agaust such lucid commentary. :-)

Nonetheless, I'd agree with Harry that's it's too long for an op-ed (just perfect for a Daily Duck post though :-).

I have a couple of questions/comments.

1. Why does affirmative action not help "to bring racist admissions boards to heel". If there were racist admissions boards, wouldn't affirmative action quotas bring them to heel?

2. You seem to admit in paragraph 8 that blacks per capita aren't as prepared as other ethnicities. In that case it wouldn't be true that "race-blind merit alone would guarantee diversity," at least at elite schools, at least from a culture/ethnicity/race point of view.

3. I'd personally lose paragraph 4. Too many words per unit impact.

4. The logic's not clear to me in paragraph 5. For example, it seems to me that being admitted to elite schools when underqualified shines a bright light on the failing schools from which the candidates come. Even if that's not so, it's not clear to me how getting rid of affirmative action will help increase the pressure on failing schools.

5. In order to keep it as short as possible, I'd consider losing the bondo metaphor (but if that's an amazingly powerful metaphor for Detroit, then maybe not).

6. It's not clear to me what the bureaucratic spoils system is. What spoils are being doled out exactly? The admission to the university? If so, why is that bureaucratic? In other words, how do the bureaucrats benefit?

I realize my questions are weak, but on the other hand, those potentially reading the op-ed probably won't have a much better grasp than me, on average.

I think I'd be more compelled if you picked just one (or maybe two) of the best point(s), focused on that, and then moved to the conclusion. If I'm a representative reader, I don't feel all that much more enlightened than when I started reading your piece. Too many concepts, too little time.

March 21, 2006 9:21 PM  
Blogger Brit said...

I was more enlightened after reading it.

But like Bret said, it's a great blog post, but too long for an editorial.

I'd reduce Para 3 to one sentence about Affirmative Action's supposed aims, if possible, and cut Para 4 completely. Possibly Para 7 could go too.

I wouldn't bother with the Constitution point either, but then I'm not American so it doesn't have the same force. It might even be your biggest gun.

Certainly it's a strong article, Duck. You tackle both the macro and micro problems, then offer some better alternatives, which makes the whole powerful enough.

March 22, 2006 3:22 AM  
Blogger Hey Skipper said...

Thanks, guys, keep it up.

At the Detroit Free Press, editorials are limited to 850 words. If I remeber the word count on this correctly, it comes in at 771. However, just because it is short enough doesn't mean it shouldn't be shorter.


1. Because college admissions boards are not racist (excluding affirmative action) -- no one is accusing them of unfairly applying admissions standards.

4. I believe the act of admission dims the light that should shine on crappy schools -- it is a classic form of moral hazard.

Your other points are excellent. I'm a little pressed for time today, but my goal is to edit the post per everyone's suggestions and see how it turns out.


I'm wondering about the Constituion point, too. While it is fundamental, I don't think people are looking at the issue in those terms. Just as a historical note, I wrote and submitted this several years ago, just before the U of M admissions case hit the Supreme Court.

The Free Press contacted me to make sure there were no rights issues with publication.

The next day the Iraq invasion started. Timing is everything.

So I decided to sit on it in the expectation a referendum would happen.

Anyway, that is a long explanation of why the Constitution para was in there then, and should probably get cut now.

March 22, 2006 4:29 AM  
Blogger Brit said...

Sorry - it's a strong article Skipper.

So much for my proof-reading skills...

March 22, 2006 6:20 AM  
Blogger Duck said...

I think you've written a very clear and persuasive essay. I would agree that for the purposes of getting it considered for the Letters to the Editor it needs to be shortened, but if you were an established editorialist I'd say it is a good length. It is all about getting and holding attention.

I'd suggest the five paragraph essay convention that I learned in colllege Effective Writing 101. The first paragraph introduces the topic and states your main argument. Paragraph's two, three and four are your three supporting facts/exhibits in order of decreasing importance. Paragraph five restates your main argument and includes some concluding warnings or predictions about the impact of matter at hand.

March 22, 2006 6:48 AM  
Blogger Duck said...


To answer #6, I'd say that as with any bureacratic system, the spoils to the bureacrat are the opportunities to keep and/or advance his career within the bureacracy. Admissions directors that play by the rules and implement affirmative action get rewarded by bringing in grants and scolarships for the university from the keepers of the spoils. Those that don't invite investigations, lawsuits and negative publicity for the university, thereby putting their continued employment at risk.

March 22, 2006 6:54 AM  
Blogger Hey Skipper said...


To clarify, this is to appear as an Op Ed piece, not a letter to the editor.


The term "bureaucratic spoils system" does not mean that the spoils go to a bureaucracy, but rather that bureaucratic means are used to provide spoils to some group.

The major beneficiaries of subsidies are sugar farmers, not the farm bureaucracy.

Consequently, the recipients of the spoils fight tooth-and-nail any move to reduce or eliminate those spoils.

Unfortunately, while that is exactly the correct term, that it is confusing to this group means I have to find a different way of saying it.

March 22, 2006 9:01 AM  
Blogger Bret said...

Hey Skipper:

Both Duck and Brit feel that your article is "strong" and "persuasive." That's great, of course, but it leads me to two questions. First, I'm wondering if they agreed with your general conclusions and perspective prior to reading the article (I suspect yes). Second, who are you writing the article for - those who already agree with you (in which case I'd follow their feedback), or those who are undecided, or those who disagree with you. My feedback would be significantly different depending on who you'd like to address.

March 22, 2006 11:00 AM  
Blogger Hey Skipper said...


The undecided and disagreeable.

March 22, 2006 1:00 PM  
Blogger Bret said...

I'd describe myself as undecided on this issue. I'd like to start round 2 of my feedback with the following description of affirmative action from wikipedia:

"Affirmative action began as corrective for past governmental and social injustices against demographic groups that have been subjected to prejudice. Such groups are characterized most commonly by race, sex, gender, or ethnicity. Affirmative action seeks to increase the representation of these demographic groups in schools, in work places, and in society in general."

The benefit (in theory) of affirmative action, is to thrust under performing minorities into situations that they would not otherwise obtain, indeed are not even qualified to obtain, so that they have new, positive experiences that will be in turn communicated and shared with other members of the minority members' community, and to begin the virtuous circle of opportunity with their children. Eventually, these minorities will perform up to their potential, and when everbody performs up to his or her potential, society as a whole will greatly benefit.

So, that's the theory, no? It may or may not be correct. I don't know, but it's a compelling theory.

If the theory is correct, then the benefits are huge, and the costs that you've identified in your article are small in comparison.

I'm thinking you could have a much greater impact on undecideds by explaining why affirmative action didn't work, or doesn't work. Show that the benefits are minimal. It seems you're already convinced of that. Please tell me why.

March 23, 2006 1:02 AM  
Blogger Hey Skipper said...


Hmmm -- good point. I'll have to work on that.

My bottom line objection is that defining diversity by skin color is practicing the very racism for which affirmative action is supposedly the corrective.

Begging the question: why does an African American from center city Detroit add any more diversity than a someone from an Appalachian mining community?

March 23, 2006 4:03 AM  
Blogger Harry Eagar said...

Here's a story from the actual trenches in actual Detroit. I got it from my brother, who was part of it.

Some years ago, Ford decided it needed to have more 'diversity' in middle management (or perhaps ultimately upper management) of its engineering sector.

It had a goodish number of black graduate engineers, so it contracted with MIT to provide at-work classes toward master's degrees.

50 students were recruited. 48 washed out.

MIT and Ford asked themselves what went wrong and concluded that the students did not know how to study. (As a dropout from a good but less prestigious engineering school, I think there's something wildly wrong about that; the 50 were GRADUATE engineers from some sort of school. Unless the school had very low standards, you had to be able to study to get out. But I am relating the story as it was told to me.)

So they took another batch of 50, taught them how to study and 48 out of 50 earned the advanced degree.

I will be back tomorrow with another real-life story about affirmative action.

March 23, 2006 10:35 AM  
Blogger Bret said...

Hey Skipper wrote: "My bottom line objection is that defining diversity by skin color..."

I thought the purpose of affirmative action is to identify and address the untapped potential in various populations. One such method of identifying such populations is by using features like skin color. Do I not have that right (as I said, it's not a topic I know much about)?

Also, the word diversity did not appear in the article you linked to. Admittedly, I'm not exactly sure how diversity and affirmative action are related.

What is the text of the ballot measure anyway?

March 23, 2006 3:40 PM  
Blogger Hey Skipper said...


Affirmative action awards race-based preferences for college admissions, government hiring, and contracting.

Here is the text of the Michigan Civil Rights Initiative.

Several years ago, California passed a similar intiative. African American representation at elite universities plummeted; they were largely replaced by Asian Americans, who, despite being a minority, and seemingly amply endowed with diversity, were not an officially aggrieved group. Therefore, affirmative action acted to suppress their numbers, despite their stellar academic credentials.

Not the least ironically to anyone with more than a room temperature IQ, though, African-American college graduation rates went up because -- here's a real shocker -- elite universities are more academically demanding, placing a premium on preparation.

March 23, 2006 5:14 PM  
Blogger Oroborous said...

...and those who disagree carry something between a whiff and outright stench of racism.

...dealing with problems rather than symptoms...

To me, those two lines were awkward.

Maybe just "those who disagree carry a whiff of racism" and "on dealing with problems rather than just symptoms", or cut the last line entirely.

March 24, 2006 5:49 AM  
Blogger Duck said...

I still think it is an excellent and persuasive essay. Some imperfections may remain, but I think that it is good enough for print. More comments later.

March 24, 2006 7:30 AM  
Blogger Bret said...

I think version 2 is a big improvement. I think the introductory and summary paragraphs are strong.

Unfortunately, the link you gave for the michigan civil rights initiative doesn't seem to work for me. Could you try again? Thanks in advance.

March 24, 2006 9:07 AM  
Blogger Harry Eagar said...

Hive mind, indeed. It's sweet as honey.

Don't over-edit, and don't listen to us so much that the piece loses its Skipper.

You're good to go.


Another true story from the affirmative action front.

In 1972, Georgia Tech wanted to do something to get the number of black students in engineering up -- from virtually zero.

With Ford Foundation money, the school invited 50 top black students (I think all boys, although at the time this was not presented to me as part of the story) to the campus in Atlanta for a recruiting/indoctrination/get acquainted visit of several days.

The 50 were found by circularizing high school principals.

When the faculty group that was to show the boys around went to the bus station, they were surprised to discover that they had 48 black kids and 2 white ones.

Inquiry showed that one of the white ones had filled in his application blank for race as 'native American.'

The other came from a school in the Alabama Black Belt with about 90% black enrollment. His (black) principal nominated him because he was a minority.

March 24, 2006 11:19 AM  
Blogger Hey Skipper said...


Try this.

March 25, 2006 3:16 AM  
Blogger Bret said...

Hey Skipper,

I agree with Harry that editing in any further input will dilute its "skipperness" and detract from the piece, so I'm gonna hold further comment for now.

Your second link worked - thanks.

After you submit it, I'd like to continue the discussion sometime, because I'll admit I'm still a bit confused by the issue.

March 25, 2006 9:28 AM  
Blogger Susan's Husband said...


I would say that the primary flaw in your argument is the idea that thrusting people in to situations for which they are not ready or trained is unlikely to result in "positive experiences" for them or anyone they work with. It is, in fact, more likely to perpetuate racial stereotypes than ameloriate them. This is precisely the point about graduation rates – what is the "positive experience" to be shared of getting in to Harvard and then flunking out?

On the other hand, my opposition to AA has been based on a simple question: "who is black?". When you can answer that for me, I will be willing to discuss affirmative action. One need only to think of who it is that has generated so much literature on that subject to see the problem.

March 25, 2006 10:25 AM  
Blogger Bret said...

Susan's Husband:

Regarding the positive experience (or lack thereof) of the recipients of the benefits of affirmative action: I don't doubt what you say is true, but I don't think the point is to provide people with positive college experiences. The bigger picture, I think, is that many of them do graduate, and in the mean time they are thrust into a wealthier, more educated culture, which has several benefits. First, the friends made and the contacts gained from a place like Harvard (where both my wife and sister went) are worth far, far more than the degree or education itself. Second, the importance of education and other values will hopefully rub off on the affirmative action beneficiaries which they'll then carry back to their communities (or otherwise provide a positive example).

That's not to say it's perfect or even efficient. But I don't doubt that there are some benefits.

Regarding your point about "who is black?" It seems to me that you're saying the affirmative action brush is too broad and therefore not particularly efficient or fair. No doubt true. But this strikes me as a case of letting the perfect be the enemy of the good. No good program will ever be without some downsides. A strict libertarian therefore concludes that no programs should ever be done. But, not being a strict libertarian, I'm willing to accept some bad with the good as long as the benefits outweigh the costs. I don't know whether that's true in this case, but neither of your points helped me evaluate that.

Can anybody address the bottom line? Are there any data which provide evidence as to the benefits and costs of affirmative action? I'm a numbers kinda guy, and while principles can easily outweigh the numbers for me, I've yet to see a truly important principle that leads me to one side of the debate or the other, so numbers it is. The closest principle that resonates with me is hey skipper's ninth admentment, but even there, exactly what "equal" means has some room for interpretation (in my opinion).

March 25, 2006 11:34 AM  
Blogger Susan's Husband said...


No, by "who is black?" I meant, can you provide a rule by which this can be determined? Affirmative action places a legal burden on an employer. Let us presume for simplicity that this is simply to achieve a caucasion / not-caucasian balance. How, in legally binding manner, does an employer count how many of his employees are caucasion? Suppose, as an employer, I simply claim that employees with names starting with the letters A-M are caucasion and N-Z are non-caucasion. Is there any legal problem with that? Would it achieve the goals of affirmative action? If not, on what legal basis could you dispute it?

This, to me, is the very heart of the problem with AA.

March 25, 2006 3:11 PM  
Blogger Bret said...

Isn't the rule to determine minority status generally written into the AA law? Isn't it usually pretty much if you say you have black ancestors within so many generations, then you're black?

Sure, people can cheat, but people can cheat at anything. That's what I mean by the perfect being the enemy of the good. It's not perfectly accurate, and people can cheat, but it doesn't strike me as a significant issue.

March 25, 2006 5:09 PM  
Blogger Susan's Husband said...

No, actually, it's not. As far as I can tell, there is no legal definition of minority status, except for American Indian (in that case, it's being a member of a government recognized tribe). But think of it – AA requires the government to create legal definitions of "race", the mark of a racially repressive regime. And who is that spent so much effort on writing on this issue? Why, the racial purity types. Is that really the tradition you want to base a government program on?

March 25, 2006 5:55 PM  
Blogger Bret said...

OK, a quick google search indicates that you're right about there not being specific definitions for what minority status is for many AA related laws and policies. I would agree that that is a significant problem.

Well, as I said in my first comment to this post, I really don't know much about AA. I guess I'm now enticed to research it a bit.

In the meantime, good luck to hey skipper and his op-ed.

March 25, 2006 6:45 PM  
Blogger Harry Eagar said...

The standard -- not a legal one any more, though it used to be -- remains in practice what it always has been: If you have any African ancestor, you are 'African.'

I was raised among those who said the precise standard was your last 400 ancestors. Never mind how that seems incompatible with an ancestry that goes by factors of 2. We're talking about crazy people.

Dawkins, in 'The Ancestor's Tale,' has quite a bit about this, notably a picture of Colin Powell and Daniel arap Moi standing together. Who's a black man?

In pre-collapse South Africa, in doubtful cases, there was a woman functionary who observed the person in question and decided, legally, to what category he belonged. (More difficult there, as the choices were among 4, not between 2.)

Among the many problems with the Michigan affirmative action law that was struck down was its contempuous provision: If you claimed to be non-white, you got 20 points (out of a required 100) on your application score. For liberals who, in other educational contexts, have been death on 'tracking,' this was a very strange situation.

That they could be on both sides of the issue just proves that what they were about was an arbitrary spoils system (or lottery, as I suggested and Skipper adopted) and not about any principal.

You may say you are more concerned about outcomes than about methods, but I'd say that when it comes to having government operate like a lottery, you're asking for trouble.

March 26, 2006 12:56 AM  
Blogger Hey Skipper said...


Thanks very much for the lottery idea -- it would never have occurred to me otherwise.

Susan's Husband:

But think of it – AA requires the government to create legal definitions of "race", the mark of a racially repressive regime.

Exactly. And further proof that irony is the universe's fundamental force.

I'll bet it has never occurred to anyone in favor of AA that it rests upon racial classification, a virtually defining characteristic of,

wait for it,

South Africa.

And that the other famous example of group quotas is Lebanon.

And we all know how well that turned out.

March 26, 2006 7:57 AM  
Blogger Bret said...

hey skipper wrote: "I'll bet it has never occurred to anyone in favor of AA that it rests upon racial classification, a virtually defining characteristic of ... South Africa."

I don't think so. The defining characteristic of South African Apartheid was that of oppression. That's completely different from the defining characteristic of AA in the United States.

Note that AA could be implemented without refering to race at all. It could classify people according to whether they are descendents from slaves who lived in the United States.

Let me ask this. If there was a way to magically increase the performance of groups performing below potential using taxes and regulation, would you be for it?

March 26, 2006 10:16 AM  
Blogger Harry Eagar said...

You didn't ask me, but I don't believe in magic.

I also don't believe in test scores. However, if that's what you're going to use to determine admissions, use the same scoring system for everyone.

The pro-affirmative action people have a number of studies out that purport to show that outcomes of students admitted to (notably, law) schools are the same for those admitted on a quota as for those who got in by merit.

I do not believe those studies at all.

No doubt in my mind that students who are admitted with substandard scores are then graded easier as they go through. You can tell that when you have to work with them as graduates.

Here's another non-racial affirmative action story from my brother.

A corporation he consults with (another auto manufacturer, I think) wanted to have its engineers get advanced training.

So the MIT faculty taped a series of courses. According to my brother, although the 'quality' of the working-stiff engineers was not as high as that of undergrads at MIT, the outcomes were the same.

In an assessment, he was told by one of the working stiffs that the reason was that, since the lectures were on tape, they could watch them over and over till they got it.

Works if the lectures are good to begin with, which was not my experience at Cow College. (And if the watchers are motivated.)

The whole affirmative action argument is bogus to the extent that it doesn't identify an educational problem first and then seek to solve it. It identifies a social problem, and instead of solving it, it jumps to the desired outcome by fiat.

No question in my mind that it would pay American society to reach out to non-participatory groups. Adding 20 points to their admission score is derisory.

Really reaching out would require money, effort, deep thinking and honesty -- all missing from the affirmative action cabal.

March 26, 2006 11:20 AM  
Blogger Bret said...

I was asking you as well and I'll take your answer as a qualified yes. So let's do some deep thinking. Unfortunately, it'll take me awhile (like more than a year) to learn enough to do deep thinking about it.

March 26, 2006 12:24 PM  
Blogger Hey Skipper said...


Note that AA could be implemented without refering to race at all. It could classify people according to whether they are descendents from slaves who lived in the United States.

Descended to what degree? Almost (but, anymore, not quite) all "Blacks" in the US are descended from slaves. As are many "whites". Hence the invidious race classification. There is some degree of whiteness that, having too much of it, makes one undiverse, and ineligible for consideration beyond merit.

But the bigger question is whether society at large owes any particular debt to those whose only relation to that awful institution is geneological.

Let me ask this. If there was a way to magically increase the performance of groups performing below potential using taxes and regulation, would you be for it?

Well, I expect if it was truly magic, it wouldn't involve taxes & regulation.

More seriously, I don't believe there is any way to alleviate group underperformance except by completely ignoring group preferences. There is no way out of that trap except through personal accountability.

The track record of the Government doing favors for African-Americans, no matter how well intentioned, has been a disaster.


I didn't used to believe in test scores, until I went to a Squadron Commanders' Conference. It was directed primarily at minority and female representation and performance in pilot training.

There is a battery of tests that officer candidates must take prior to joining. In some respects, they are like SATs; in others, more like an IQ test -- lots of mechanical reasoning and shape matching.

I had always assumed that, within a pretty wide range of scores, these tests were pretty much worthless.


When grouping pilot training qualifying scores in ten bands, then matching performance in training against the scores, the trend was remarkable. In the lowest decile admitted to pilot training, the failure rate was (IIRC) roughly six times that of the highest decile.

Clearly, the results showed that testing was a valid admission standard, and that lowering the requirement any further would result in unsustainable costs in terms of wasted effort due to the inevitable increase in washout rate.

(BTW -- the test was under some criticism by some groups because, according to them, the results were insufficiently diverse.)

Because of the largely effective screening, I have very few AA tales to tell.

At one point, I did have the distinctly unpleasant task of washing out two African-American students.

While I am certain they were intrinsically capable of succeeding, they came to pilot training with weak reading, and practically absent studying, skills.

They could only have gotten to my squadron via the AA escape clause, which every step along the way allowed skin color to provide the education system an excuse to avoid setting standards.

March 27, 2006 4:59 AM  
Blogger Harry Eagar said...


Perhaps I should hone my statement to something like: 'I don't believe in general test scores as predictors of general ability.'

That's just a restatement of Gould's 'Mismeasure of Man.'

I do accept, at least for starters, that there exists 'intelligence,' and therefore there must be a way to measure it.

Or, following Howard Gardner's 'Frames of Mind,' 'intelligences.'

I don't think I want to be operated on by a brain surgeon who never had to pass a skills test -- in brain surgery. I don't much care how he did on analogies.

When I was washing out of engineering, I went to the guidance office and took the Minnesota Multiphasic and some other tests. One put me in the bottom quintile of college students in what you call 'shape matching.'

Yet I scored really high on SAT type tests. All that proved, really, was that I was a bookworm.

March 27, 2006 11:05 AM  
Blogger Brit said...

Some very intelligent people are no good at written examinations, for reasons varying from nervousness to dyslexia. And conversely, some not very bright people are very good at passing written examinations, because they've been well coached for a particular test.

Every year in Britain the national average A-level (18 year olds) and GCSE (16 year olds) exam results get better, and we have the tedious perennial debate: are exams getting easier, or are British kids getting smarter?

Since the latter seems unlikely, most people think the former. But there's a far more likely third answer: the teachers are getting better at coaching students to pass these tests, precisely because the tests are very similar every year.

Teachers can take past exam papers, and coach students to write good essays and answers to them, knowing that this year's questions will be pretty much the same.

So as long as you have a moderate intelligence and apply yourself, if you go to a decent school you'll do a lot better in exams than a more intelligent kid at a bad school.

That's why the Oxbridge universities still prefer interviews with prospective students.

However, since interviews are impractical when dealing with vast numbers and limited time, and since most intelligent people are better at exams than most stupid people, written tests are a necessary evil.

March 28, 2006 12:24 AM  
Blogger Harry Eagar said...


American high school students don't have to write anything. Just fill in circles on a 'machinable' answer form.

Actually, I think the latest revisions do require a little bit of writing, but the length is not much longer than the average comment at Daily Duck.

March 28, 2006 7:58 AM  
Blogger Harry Eagar said...

Leave us hope the editorial op-ed page is more balanced than that.

Op-ed pages are not supposed to be just more of what is on the ed page.

March 31, 2006 10:52 PM  

Post a Comment

Links to this post:

Create a Link

<< Home