Hacker News Reader

Comment by gxs | Hacker News Reader

I have always wondered what exactly is going on in society:

People in the US, more so than other countries, have a plausible chance of going to college, even if it takes going in to massive debt.

Then, they will go on to find a mate who is likely also college educated, and in a similar income bracket.

Over time, doesn't this mean that we're breeding (for lack of a better term) a class of people who are a) capable of being educated and b) seek out education? Doesn't this only widen the achievement gap?

Not proposing a solution, but I suspect if we were to take this stuff seriously, we could make serious inroads to addressing inequality.

On the other hand, as a country we rightfully (IMO) believe everyone is equal and work hard to provide opportunity for everyone.

What's the right move here?

View on HN

The Bell Curve (1994) talked about the intellectual stratification that is happening in society. The author and resulting research was called "racist" and has been shouted down time and time again but much of what you say is discussed in the book. He shows correlation between intelligence and [income, parent's socioeconomic status (SES), job security/unemployment, educational attainment, and tons of other things]. The data is sporadic pre-WW2 but complete and well-validated afterwards. He avoids drawing conclusions or prescribing solutions.

All of it is interesting and some is compelling, but the book is generally forbidden to be discussed in "polite" company.

The interesting thing about the book.. to reduce the "racist" attacks, the author only looked at white people throughout the bulk of the book.

Wow, fascinating, thanks for the book reference.

Reading the wikipedia article made me uncomfortable - I doubt a book like that would ever be written today.

    > the book is generally forbidden to be discussed in
    > "polite" company

The problem with studies like this - and also ones that for example attempt to show race-linked differences in "positive factors" - is that they're pretty useless for anything other than the uneducated using as a blunt discrimination tool.

Let's say you live in a society where there are blue and green people. A study shows a genetic basis for a difference - on average across the whole population - where blue people are 10% less intelligent and 10% more violent.

What's next? Are you justified in preferring to hire green people? No. Are you justified in having greens-only policies? Of course not, that's not how averages work. Is there any basis for any kind of segregation? Any basis for any color-based policies? No.

But you can be sure that there will be green people trying to use these studies to show themselves as superior, as a basis for discrimination, for popularist politics, and all the other associated bullshit.

It's precisely because there's almost no practical use for studies like these, but they are absolutely pounced upon by racists and opportunists, that they're 'generally forbidden to be discussed in "polite" company'.

Ignoring reality is the worst possible option. Regarding your assertion, judging an individual by their collective group average is wrong. I'm a big proponent of research in this area - because I want to change it. For e.g. Black Americans have an IQ, on average, 15 points lower than white Americans. I want more research on this area, so it can be improved. Calling people racists (without evidence) shuts down any meaningful conversation, and hurts those who need the most help.

    > I want more research on this area, so it can be
    > improved

Somehow it sounds more obviously ridiculous when you rephrase it and start looking for "solutions" to improving the intelligence of all white people.

Small differences between very large, heterogenous populations gives you virtually no predictive power about individuals or how to help them, but do facilitate subtle and overt racism, however well-meaning it might be.

On what evidence can you say that more research won't/can't help?

[1 SD is not a 'small difference'. (2 SD's if you compare to Ashkenazic Jews).]

    > On what evidence can you say that more research
    > won't/can't help

On what basis can you say it will? These studies have known drawbacks in that they serve as a rallying point for all sorts of discrimination masquerading as high-minded discourse, and we're yet to see a benefit so far, so the burden of proof here is on you.

At least part of the problem here is given two children / job-applicants / defendants / whatever, who are absolutely identical in every way except for a specific cosmetic phenotype, what possible justification is there for treating them differently based on correlation between that phenotype and some other characteristic over a large population?

You are all over this thread putting words in people's mouths instead of discussing what they are actually saying. You are even painting an entire class of study with a broad brush in a guilt-by-association way. Are you sure that's actually helping your case and the conversation here?

What's an example of words I'm putting in someone's mouth? Perhaps you could reply specifically to a case where you believe I'm doing that?

> These studies have known drawbacks in that they serve as a rallying point for all sorts of discrimination masquerading as high-minded discourse, and we're yet to see a benefit so far, so the burden of proof here is on you.

You're arguing for deliberate ignorance because reality is uncomfortable.

> At least part of the problem here is given two children / job-applicants / defendants / whatever, who are absolutely identical in every way except for a specific cosmetic phenotype, what possible justification is there for treating them differently based on correlation between that phenotype and some other characteristic over a large population?

If two people are identical in every observable way, and that single observable way indicates a strong difference between them, why wouldn't one differentiate based on that single observation? They are, after all, indifferentiable in every other way.

Indeed, mathematically speaking if one ignores the single useful datum you posit, then one will make worse decisions: if you have no other data and just flip a coin between greens and blues in your thought experiment a few posts back, then you will choose worse than if you bias your coin towards greens. That's just a fact; whether you choose to ignore it based on the discomfort it causes you is up to you.

A 10% difference between the populations they're chosen from does not indicate a strong difference between them.

The usefulness is if you see a deficiency in a student (or anyone) you can properly understand the reason for it.

When you understand the reason you have a chance at actually making things better.

I.e. use the data after seeing the performance of someone, rather than using it to predict their performance.

    > When you understand the reason

This right here is the problem.

Armed with a small amount of data about information over a population on average gives you absolutely no information about any specific case, and the green-skinned child gets special treatment to bring them up to scratch, where the blue-skinned child gets written off as "she's achieving less well because blue-skinned children do".

That's the problem, and that's why people in "polite societies" ignore these studies.

You realize there is not a small amount of data regarding racial IQ differences right? It is (was) one of the most studied social science fields before it fell out of fashion.

You also seem to want to reduce something to skin-color that has been scientifically identified as more than skin color.

Please could up explain the point you're trying to make? The first sentence suggests that you think my point is lack of evidence is a problem, the second suggests you think my argument relies on one specific factor rather than being generalisable to all.

Sure, there are two points:

1) Your point is based on ignoring often repeated and consistent (over decades) science; and

2) I think your argument generalizes something (to skin color) that is very specific (I.Q.)

I will go further and say that I think you are being disingenuous about reducing the problems that affect different races to simply skin color.

1) I'm not ignoring it, I'm saying it's unimportant for any practical purpose, and the only people who seem to get excited about it are people with an existing axe to grind, which is why it's not generally discussed in the op's "polite" society.

2) The article this conversation is attached to is specifically genetics related.

1) It is very important as it has been found to directly relate to educational achievement and later SES. The fact that "polite" society turns a blind eye does not make have less of an impact.

2) Yes, but the comment you were replying to (and made) was not.

> Armed with a small amount of data about information over a population on average gives you absolutely no information about any specific case

That's akin to saying that being armed with all of medical science's knowledge about appendicitis gives one absolutely no information about any specific case. There's a valuable point buried deeply there (every case is different), but the statement itself is just plain wrong. A physician or surgeon must have reasons to treat a specific instance of appendicitis differently than the statistically expected range of instances; a person would have reasons to treat a random green or blue person differently than the statistically expected range of persons.

> the green-skinned child gets special treatment to bring them up to scratch, where the blue-skinned child gets written off as "she's achieving less well because blue-skinned children do".

Economically speaking, investing resources in bringing green children up to scratch has a higher rate of return than doing the same for blue children. That really does mean that spending resources on the latter is wasting them on average (although certainly not in specific cases, where it really is a good investment). So, do you impoverish the world by wasting resources or do you enrich it by prudently investing them?

However let's say the figure is 50%. Does the argument still follow? At what % would we inevitably be forced to pay attention? For instance, would anyone want to increase the chance of a violent encounter occurring in their workplace. Fortunately we can relax because these supposed assumptions currently belong in fairy land.

'Popularist'. Isn't it better to spell out what is meant here rather than use a word often by elites (self selected smaller powerful groups) to refer to any democratic decision of which the elite does not approve? It's one of those rolling stone expressions that saves its users time and effort by carrying with it a boat-load of subtext.

    > 'Popularist'. Isn't it better to spell out what is
    > meant here

You're welcome to substitute the word "facile" if you've imbued "popularist" with too much meaning.

That's much better! Of course I don't matter a jot but many folk who do (some EU leaders for instance) evidently believe 'populist' (as a word of derision - your implication, surely?) is indeed imbued with rich meaning of the kind they would prefer not to to spell out since their argument would then fall to bits.

    > However let's say the figure is 50%.

It's not.

   > Does the argument still follow?

Yes. Given two people with similar academic achievements and no history of violence, discriminating between them is ridiculous. Treating individuals different based solely on their skin colour would still have no scientific or logical basis, as you'll still have many many stupid and violent green skinners, and peaceful and intelligent blue skinners.

All that the "studies" will do is discourage natural bigots from looking at the traits of an individual.

    > At what % would we inevitably be forced to pay
    > attention?

At the point where you could reliably extrapolate about a given individual based on their skin colour more accurately than you could based on their personal history to date.

> It's precisely because there's almost no practical use for studies like these (...)

Let's say that green people have a higher rate of suicide attempts when they have mental illness X. This is very useful knowledge -- this means that one should treat X in green people more aggresively, because it's more dangerous for them.

Let's say that green people are more likely to suffer from alcoholism. This is very important for green people themselves (they might prefer being teetolaters for that reason) and for organisations that try to prevent alcoholism -- they should target green people more in order to be more effective at decreasing the number of alcoholics.

I can go on and provide such examples. Do you consider that in these cases there is still almost no practical use, that these examples are thought up and in real life you don't get such or something else?

    > This is very useful 
    > knowledge -- this means that
    > one should treat X in green
    > people more aggresively, 
    > because it's more dangerous
    > for them.

No you shouldn't, you should asses each patient individually, especially if - as per say qualification for a job - this fact is discernible by actually examining the person. Otherwise you will let a small 10% increase in likelihood for one group blind you to the needs of another.

> No you shouldn't, you should asses each patient individually

Assessment means making observations; making observations means observing; observing means actually opening your eyes to the facts as they are. You'd be killing your patients if you ignored the fact that greens commit suicide more often on average.

In that scenario, assuming a 50/50 population split of blue/green people, should you aim for having 50% of each employed at your company or are you justified in hiring fewer blue employees?

For a 10% difference? Unless you are employing a huge workforce where intelligence and peacefulness are the key features, I think you should be aiming for 50%. There are vanishingly few professions where you'll notice a 10% difference in IQ.

Thank you for doing yeoman's work here.

> A study shows a genetic basis for a difference - on average across the whole population - where blue people are 10% less intelligent and 10% more violent.

> Are you justified in preferring to hire green people? No.

Why the heck wouldn't you be justified in preferring to hire people who are, on average, more intelligent and less violent?

> Are you justified in having greens-only policies? Of course not, that's not how averages work.

I agree — in the scenario, any individual blue might be preferable to the great mass of greens.

> Is there any basis for any kind of segregation? Any basis for any color-based policies? No … It's precisely because there's almost no practical use for studies like these

There's no practical use that you care to put into practice. What you're arguing for is categories of 'forbidden knowledge': stuff that everyone agrees to ignore, because its truth is too discomforting. That seems a bit childish to me: one of the hallmarks of maturity is coming to terms with uncomfortable knowledge.

Besides, here's one practical use for such a study: in your example, if blues were a much higher proportion of the prison population than they are of society at large, knowing that they are less intelligent and more violent on average is information which can help one determine if that is fair or not.

    > Why the heck wouldn't you be
    > justified in preferring to
    > hire people who are, on 
    > average, more intelligent
    > and less violent?

Because that knowledge tells you absolutely nothing about any given individual.

> Because that knowledge tells you absolutely nothing about any given individual.

You can't know the heart of any man: all you can do is make predictions about how he will behave under various circumstances, based on the observations you have made of his behaviour so far. To ignore an observation with strong statistical power because you don't like it is simply unscientific.

    > based on the observations
    > you have made of his 
    > behaviour so far

That's what interviews and resumes are for. I would be interested to know how you would practically hire for race while not being hit by the halo effect for one race on one hand, and not missing strong candidates through racism on the other.

    > What you're arguing for is
    > categories of 'forbidden
    > knowledge'

No, I'm expressly calling attempting to make useful decisions on individuals based on small differences in wider populations illogical and non-rational. It would not be difficult to model this mathematically.

You should really read my whole comment instead of jumping on what you guess is in the book and what "study" you think the author performs.

As noted in my comment, the author looks exclusively at white people. More importantly, he looks at a variety of long term (aka multi-decade) studies from a variety of US sources and analyzes that data.

But if you want to keep arguing straw men, please feel free. I won't interrupt further.

I think you'll find that birthrates are negatively correlated with education level, so globally the reverse is happening (with regards to your breeding theory).

But more generally, the reason why evolution works is because organisms are not equal (mutation and crossover are random). I don't expect that this will stop any time soon.

IIRC, birthrate (in the developing world?) is correlated with female empowerment. Small stuff like when females have educational and economic opportunities, are not utterly dependent on their males counterparts, have basic humans, etc.

At the speed science and technology are progressing, Darwinian evolution is more or less irrelevant as far as humans are concerned. We'll have IQ enhancing drugs, invented hard AI to take care of us or cured age death long before random DNA mutations or "breeding" adds or subtracts a few IQ points to our descendants. There are a lot of things we should worry about but genetics of future humans isn't one of them in my opinion.

Equal opportunity, not equal outcome or talent.

I agree, although I think 'Equality of process' is more accurate than 'opportunity'.

Yes, it is a well studied phenomenon called assortative mating.

Reply on news.ycombinator.com