The Bell Curve

[_EAllusion]

I made a couple of minor corrections so that it is now completely true. Thanks for bringing this sloppiness on my part to my attention.

You might need some further corrections.

Chapter 15 - The Demography of Intelligence - is also about race and IQ. Or rather, that chapter is largely about their concern that the wrong sorts of people are reproducing too much and the right sort of people too little followed by a similar discussion on concerns with immigration lowering the IQ of the population. Ethnicity and IQ are discussed a lot in this chapter in that context. This part of the discussion relies on the congenital nature of IQ as it relates to race and a substantial focus is on their belief/prediction in further divergence of IQ between racial groups due to differences in reproductive rates. It does not state, but heavily implies that low IQ racial groups elevated fertility in comparison to whites is a problem worth worrying about. It sort walks you the reader right up to that inference, but doesn't say it out loud. Quoting that discussion basically would require me to quote most of the chapter. Given that you read it, do you believe the discussion therein is also on solid ground that's only been further established as time has gone on? If so, what research would you point me to for that position?

Normally, when a reader and the author agree on what a book is about, it is a sign of clear writing and reading comprehension. But in this case, it sounds like you are accusing me of plagerism. If so, please do two things. First, send my words and what Charles Murray said about the book himself to Lemmie and kindly ask her to analyze it and share her conclusions. Second, go to hell.

Yeah, I was coyly accusing you of copying a comment Charles Murray made about his own book in defense against its critics that is obviously false. I did this because I respect you enough as a reader to realize how this comment is error in several ways, thus leading me to think you were repeating defenses of it rather than generating the same incorrect reading yourself. The "just chapter 13" line is otherwise really strange when you have the book sitting in front of you contradicting that point. It's also wrong to say that The Bell Curve, outside of the racial context, is down-the-line rock-solid, uncontroversial science, but you put yourself in a position to argue that even a lot of its incendiary commentary on race is well-established. This isn't true, and further should make a reader of you wonder where you got that notion from.

[_EAllusion]

The book clearly and repeatedly indicates that IQ is malleable--just not as malleable as many people seem to think. They do talk a lot about how if you study for an IQ test you can improve the score, but they wonder how much of that is actually changing the underlying g. But beyond that, they repeatedly say that getting an education does in fact increase your g. In fact, going to school for 12 years (as opposed to not going to school at all), will improve your g a lot.

But that's not at all relevant to the question I'm bringing up. I'm not proposing that you advocate for shortening the length of public education from 13 years to 5. I'm talking about equalizing differences in public educational systems so the more expensive ones resemble the cheapest ones because the actual variance in quality of education is proposed to matter very little to cognitive ability as measured by aptitudes generally learned around high school age. Their reasoning, helpfully quoted by you, implies that variance in education has little impact on this outcome.

The statistical model your book criticized doesn't take quality of education (both in the community and in the schools) into account, but that is only because of limitations in the data.

The limitations in the data necessarily should temper what we can say about the effects of education on IQ.

The book itself talks a lot about these other factors and suggests that some of the best levers we have for increasing IQ is adopting kids from bad enviornments into good ones and increasing nutrition.

Outside of adoption at birth, it also downplays changing environmental contexts, including school nutritional programs, as compared to what other research generally indicates. The Bell Curve was infamously used as a cudgel against Head Start programs for that reason. It has a whole section on the relative futility of Head Start and similar programs (disputed by other researchers), in fact. My comment isn't so narrow and is about the broader ecosystem of education a person exists in during their developmental years. It starts in infancy, but I do not think research shows malleability is lost at that point. I also meant that if you are in the 3rd grade and have parents who are interested in reading, that's helpful to your school performance, which likely has downstream effects on how well you test on academic ability when you are a teenager. Likewise, quality of teaching is hard to capture with this kind of dataset, especially because the same teachers provide different levels of quality to different students.

Are you suggesting that there is robust data that equalizing the amount or objective quality of schooling can be counted on to signficantly equalize cognitive ability?

For now, I'm suggesting that if you think this, remaining agnostic on the point for purposes of discussion, then do you naturally accept the logical conclusion that we should reduce premium spending on education quality since it matters very little? If not, then why not? So far, you've reacted to this point with a mixture of offense and red herring, but it's a natural question. What originally made me ask is this conclusion, while right in the wheelhouse of a libertarian like Charles Murray, would be a radical departure from your normal politics. The book in general is pessimistic about all social uplift policy. But when it comes to education, I wanted to know if you followed through on the logic and were ready to support significant cuts to the more generously funded education systems in the country, since they are quite bad in producing differential outcomes against their cheaper counterparts as argued by the Bell Curve.

Murray would answer your question by saying that since education does improve IQ, we should spend money on it. He would further say that school is about more than improving IQ, and that those other factors make spending money on it even more worthwhile.

You're still not getting it. I thought I used a more direct question that got the point home. Why not fund Minnesota public schools like we do West Virginia's since variance in quality matter so little? It is not Murray's position that any improvement in IQ, no matter how marginally insignificant, justifies any level of spending. Murray doesn't think like that at all. He does argue that schools also involve socialization and other non-academic benefits, but he does not say those justify any level of spending you might desire for public schools. If the benefit of schools is *that* why not more cheaply redirect our resources to that than spending such a large premium on pedagogical expertise in teaching students the knowledge and mental tasks tested by the ASVAB?

Given that that is his actual position, am I evil for pretty-much agreeing with him?

[/quote]
You agree with him less than you think. Murray is rather infamous for advocating defunding or cutting many social programs on the argument they are not cost-justified in their benefits. You're are hanging on to the point that he isn't advocating ending education, but that's not what is under discussion. Further, it doesn't even matter what Murray thinks as we are talking about the implication of Murray's arguments. If it is true, as argued in the Bell Curve, that quality of education matters very little to core aptitude, then why spend so much on quality of education? You haven't begun to answer this at this point, and saying you agree with Murray sheds no light.

[_Analytics]

Hi Analytics,

Thanks for the additional information in responding to my post. The only item I remain curious about regarding the book is regarding the question of ongoing evolutionary impacts as they are dealt with by Murray and his coauthor.

I ask this because it still seems like a "so what?" book to me insofar as if the book focuses on IQ, data that points to variability in IQ that can be mapped onto ethnic groups in the US (which historically could also be defined as much by cultural as by genes given segregation, etc., etc., etc...), notes that IQ is malleable but not infinitely so such that finding an intelligent mate will improve one's offspring's chances of success in life and therefore one's genes' success, and that whatever degree one assigns to it all agree quality of education matters...again, what exactly is Murray's call to action? Be real but be nice means...?

I'm reading the material you've posted and, at least from a historic perspective, could go so far as to read it mainly providing a good reason to consider oppressive behaviors towards populations as having long term consequences that take effort and time to overcome because there are perpetuated negative impacts that don't go away with the birth of a generation that is not overtly being oppressed. Is that part of this? Whose to say this isn't actually an argument for affirmative action if one steps back and accepts the proposed reading of the data with an eye of getting away from the racist aspects of potential conclusions one could draw to the other conclusions that lean more on acknowledging evolutionary pressures and advantages at the level we're talking about aren't exactly the type of change one would assign to genetic variability that starts to get into Markk territory. If anything and were it to be taken at face value as presented, it starts to read like one of the worst indictments of colonialism possible...

Anyway, I'm still not inclined to read the book but this is starting to pique my curiosity enough I may get to it eventually.

The big "so what" have to do with a few things. First, what should our realistic expectations be for how society ought to be? Second, what kind of social policies should we implement to get from here to there? Beyond that, it is about understanding the sociological implications of IQ in our changing world--these points have cause of concern and no clear anawers.

Here is a personal example. I am an actuary, and something like 99.8% of Fellows of the Society of Actuaries (FSAs) are white or Asian. A year or two ago, the president of the SOA lamented this in a meeting and groped for a way we could become more diversfied.

Here is the thing. To become an FSA, all you need to do is pass a series of 10 tests. That is it. If you pass the tests and get your FSA, you are pretty much guaranteed a job with a 6-figure income that is consistently ranked as one of the best jobs in America. Geographically, actuarial jobs are all over, but are somewhat concentrated in northern, liberal cities with black populations like Hartford, New York, and Chicago. As far as I know, there is no history of institutional racism in the SOA or actuarial employers.

The exams are relatively inexpensive. They are largely created, administered, and graded by volunteers. And there is no admissions process--if you want to take an exam, all you have to do is register for it on the Internet and pay the test-by-test fee. The tests are then anonymously graded by people who haven't even met at least 99.9% of the candidates, but just in case the graders are not provided with any information about who wrote the paper they are grading.

Getting your FSA is arguably as valuable as an MBA from a prestigious business school, but getting it is practically free by comparison and wide open for anybody and everybody to pursue who wants to.

Here is the kicker. The exams are hard. It is difficult to describe how hard they are. Imagine taking the bar exam. 10 times on 10 different topics. It might be something like that.

So given all that, what should we do to get more black members? Should we follow what colleges have done with affirmative action and lower the bar if you happen to be black? We'd get more black members, sure. But they would forever have an invisible astrisk on their credentials that says they only passed the dumbed-down exams made just for blacks. Would anyone really want that?

I get it that race is merely a social construct and that "blacks" have imense and totally unfair and undeserved disadvantages due to risidual racism and related disadvantages. I'd love nothing more than for all of that to go away and for it to be conclusively proven that as a group, blacks are as intrinsically intelligent as whites. And I'm not saying that they aren't as intrinsically intelligent as whites.

So given all that, what should the SOA do about the racial imbalance in its membership? Murray would argue that we should be especially welcoming to blacks when recruiting and promoting the profession, and that if we are hiring and two candidates are basically equal, give the job offer to the black. Otherwise, we should be colorblind and treat everyone with equal dignity and respect, irrespective of skin color.

It seems to me that is what we've been doing, but all that has done is led us to where we are now. According to Murray's thinking, we should treat blacks and whites who don't pass exams the same, just as we should treat blacks and whites who do pass the exams the same. We should become colorblind that way.

I have strong PC sensibilities that more should be done to help blacks and make up for the head winds they face. But with regards to the problem that few blacks pass actuarial exams, what can be done? What should be done?

[_Analytics]

...I'm talking about equalizing differences in public educational systems so the more expensive ones resemble the cheapest ones because the actual variance in quality of education is proposed to matter very little to cognitive ability as measured by aptitudes generally learned around high school age....

Believe it or not, I get your point, and see how in a way, what you are saying naturally flows from some of the things that Murray says. The thing is, I read him in a more subtle way. What I hear him saying is this: Education matters. The amount of time studying matters, and the quality of education matters. It all matters a lot. Therefore, it should all be well funded. Nevertheless, we shouldn't expect that equalizing the quality and quantity of education will do a lot to equalize the results.

You seem to be saying that if Murray is right and we can't expect equalizing quality of education during school hours to do very much to equalize educational outcomes, we should race to the bottom and cut all school funding to the lowest common denominator. B doesn't follow from A (although B might follow from a misunderstanding of A).

You infer certain things from the data Murray presents. He infers something subtly different--at least in this book he claims to infer something different. My inference is closer to what he claims to infer.

Murray does declare that in terms of raising IQ, Head Start has been a failure--it doesn't raise kids' IQ to a significant degree. But he also says that it is worthwhile in terms of giving kids a good meal and a safe place to be for a few hours a day. Is there convincing research that proves Head Start does permanantly raise IQ by a significant degree?

[_EAllusion]

Murray's thinking goes deeper than that, though. Murray also thinks the reason that black people, as a group, are less inclined to do well on a test to become an actuary is at least part genetic. This is significant, because that's intrinsic to being black, not anything about the historical oppression blacks have faced and continue to face. The sources and arguments used for this idea would be humorous if the subject wasn't so ugly, which is quite frustrating when coupled with how he brushes aside widespread racism both in its current and historical effects a hypothesis for black/white IQ gaps. In fact, in Murray and Herstein's largely throwaway line about being agnostic on the nature / nurture mix of ethnic differences in IQ they explicitly state they don't think it matters, even though it matters a great deal. Then they write lots of material where it matters.

Murray wants black people to be less dumb - he talks about how concerned he is about how stupid black people are as a group - but he's also consistently down on uplift and quite dismissive of traditionally proposed causes of racial IQ gaps as explanatory. The overall thesis is that black people are poorer than whites, found more frequently in jails, and so on mostly or in notable part because they are stupider than whites with those outcomes being a natural on average consequence of being stupid. While it would be nice to fix this, there's no reason to think a fix is on the horizon because intelligence is relatively immutable and we should be aware that it's highly likely to at least be partially a hereditary issue. This is a serious topic that, while taboo, he wants more focus on. That the the idea that blacks might be genetically superior to whites in predisposition for intellect if not for environmental factors is too risible to consider. Best case scenario is that it is only a little hereditary. The test results of blacks in apartheid South Africa prove it.

In the Bell Curve, I believe affirmative action first enters by implication in a discussion on how blacks, when you control for IQ, are actually over represented in high skill jobs. It's not initially stated, but the reader is led to think by later commentary that this is a consequence of society "going easy" on blacks. This is part of the works' repeated, "we're not saying, but we're saying..." style of rhetoric.

Affirmative action is viewed as bad for the classic arguments that Analytics brings up, though with a healthy dismissal of the main justification for Affirmative action being a counterweight to ongoing discrimination that can be overcome with opportunity.

So what do you do? If you think that testing gaps that are predictive of desirable outcomes are a consequence the legacy of hundreds of years of white supremacy in America that can be gradually unwound, that's going to produce a different answer than if you think testing gaps are baked in to relatively immutable factors. Murray and Hernstein are quite dismissive of racial biases in IQ testing itself, which obviously is a major area of dispute within the field. If you think cultural test bias explains some of the gap, and that same cultural bias also relates to different social outcomes such as employer discrimination, your expectations for what can be are going to be quite different than if you think it has to do with how black brains are, to an extent, less well-suited for high level math on average.

[_EAllusion]

Since I work with intellectual disability for a living, I might be overestimating how intuitive it is for people to understand what it means to be of borderline intellect or at the borderline of IQ that is predictive of true intellectual disability. It is an extremely provocative claim that the mean IQ in Africa is around the low 70's and that should be regarded as a valid assessment of underlying intelligence. It's even extremely provocative to say the same of mid 80's IQ of African Americans. Do you understand how cognitively impoverished that is?

To put it in perspective, if you have an IQ in that range and it is an accurate reflection of your cognitive ability, I probably have a good shot at getting you government funded services including caretakers to help you function in your life at a basic level. I assessed a woman from Liberia recently who had a recent neuropsych with an estimated IQ in the high 70's to low 80's (after having an estimated IQ of around 50 when she first moved to the US at the age of 12). That low IQ and the deficit in intelligence it represented was part of the case for her need for a range of services I was there to assess.

Subsaharan Africa has over a billion people where the Bell Curve estimates mean IQ is near the range of intellectual disability. The implication here is obvious. The primary source Hernstein and Murray relies on believes that "Negroids" have never contributed anything meaningful to civilization and left to their own devices are on average unable to function much above the level of a hunter-gatherer society. As it happens, that source they rely on and describe as a foremost expert in race and IQ is a literal white supremacist who advocates for a white ethnostate inside of America and uses methodology to support his views that ought to be regarded as academic malpractice. But even if that weren't so, we should be able to grapple with the implication of mean IQ being that low and it actually being a valid measure of underlying intelligence. It's a big deal. It's wrong to gloss over it. Either it is justified with enormously important implications or it is not and it is irresponsible to the point of flat out racist to argue it.

[_honorentheos]

Murray's thinking goes deeper than that, though. Murray also thinks the reason that black people, as a group, are less inclined to do well on a test to become an actuary is at least part genetic. This is significant, because that's intrinsic to being black, not anything about the historical oppression blacks have faced and continue to face. The sources and arguments used for this idea would be humorous if the subject wasn't so ugly, which is quite frustrating when coupled with how he brushes aside widespread racism both in its current and historical effects a hypothesis for black/white IQ gaps. In fact, in Murray and Herstein's largely throwaway line about being agnostic on the nature / nurture mix of ethnic differences in IQ they explicitly state they don't think it matters, even though it matters a great deal. Then they write lots of material where it matters.

Murray wants black people to be less dumb - he talks about how concerned he is about how stupid black people are as a group - but he's also consistently down on uplift and quite dismissive of traditionally proposed causes of racial IQ gaps as explanatory. The overall thesis is that black people are poorer than whites, found more frequently in jails, and so on mostly because they are stupider than whites with those outcomes being a natural on average consequence of being stupid. While it would be nice to fix this, there's no reason to think a fix is on the horizon because intelligence is relatively immutable and we should be aware that it's highly likely to at least be partially a hereditary issue. This is a serious topic that, while taboo, he wants more focus on. That the the idea that blacks might be genetically superior to whites in predisposition for intellect if not for environmental factors is too risible to consider. Best case scenario is that it is only a little hereditary. The test results of blacks in apartheid South Africa prove it.

In the Bell Curve, I believe affirmative action first enters by implication in a discussion on how blacks, when you control for IQ, are actually over represented in high skill jobs. It's not initially stated, but the reader is led to think by later commentary that this is a consequence of society "going easy" on blacks.

Affirmative action is viewed as bad for the classic arguments that Analytics brings up, though with a healthy dismissal of the main justification for Affirmative action being a counterweight to ongoing discrimination that can be overcome with opportunity.

So what do you do? If you think that testing gaps that are predictive of desirable outcomes are a consequence the legacy of hundreds of years of white supremacy in America that can be gradually unwound, that's going to produce a different answer than if you think testing gaps are baked in to relatively immutable factors. Murray and Hernstein are quite dismissive of racial biases in IQ testing itself, which obviously is a major area of dispute within the field. If you think cultural test bias explains some of the gap, and that same cultural bias also relates to different social outcomes such as employer discrimination, your expectations for what can be are going to be quite different than if you think it has to do with how black brains are, to an extent, less well-suited for high level math on average.

I thought I'd apply the principle of greatest charity to the way the information from the book was being presented and see where it took me. Without having read the book or having any more familiarity with it's methodology or data than what's been shared it seemed the most fair approach to try and see where Analytics might be coming from in attempting to say the book was misunderstood.

The result was that if we assume there is some form of genetics-based variation between people of different ethnic backgrounds on intelligence test performance one is left to ask what would be behind it and what it means going forward if anything. And the result seemed to be at best - at best - the claim ought to be that centuries of racial animus and cultural influences can't be erased with a wish that everyone would move on from the past because the consequences went beyond merely questions of access and socio-economic status. The effects will take time and effort to unwind even if we could wipe away all other factors. That's a lot to assume but going with it didn't get me to the place it seemed Analytics noted is where Murray takes his readers.

To come down in any other place would seem to put someone on the spectrum that Markk planted a flag at one end of rather boldly. So, accepting the data for the sake of understanding, it seems to be odd that the claim is the book favors not wanting to say blacks are less intelligent than whites - and stressing no one is saying blacks are less intelligent - but let's not pretend that wink wink wink isn't true so we need to be realistic.

Ok.

Anyway, the reality is there are dozens of career paths where whites are seriously over-represented and most aren't hidden behind the walls of sounding boring as hell combined with a battery of testing and needing to know a guy to get one in the door to apprentice at it to be most likely to pass the test which is exactly the sort of gatekeeping that leads to a lack of diversity. At some level that last argument left me much less charitable than any of the discussion about Murray because it can't be attributed to either facts or an attempt to understand a researcher's argument. It's simply a blind spot to the much-reported on diversity issue in American career paths that cross the spectrum on education needed, pay, and even being blue collar v. white collar.

I don't know. It does look like it's shaping out to be people finding cover for racism lite.

[_EAllusion]

Education matters. The amount of time studying matters, and the quality of education matters. It all matters a lot. Therefore, it should all be well funded.

That's a wishful reading. The relative difference in quality of education does not matter all that much according to the material you are supporting. I already quoted you quoting it saying exactly that, so I'm not sure what else can be done to point this out. The argument is more extensive in the book about quality of education, narrowly construed, having minimal effect. This isn't a reason to end schooling in general, because the relative difference doesn't involve schools vs. no schools. It is a reason to care less about having "good schools" as opposed to "bad schools" as they actually exist and therefore gives us financial incentive to cut costs. Specifically, things like teacher credentials, class size, class materials, etc. don't matter enough to justify spending for the expensive versions. Education, in this sense, does not "matter a lot." It matters very little according to your preferred source. High performers will still perform high and low performers will still perform low on average for the most part. If high performers didn't perform well in bad schools or low performers performed well in good schools, you'd have a point, but as it stands, why not race to the bottom? The top seems like a waste. If you think spending a fortune chasing almost no improvement is worth it, remember that resources are limited and there are other things the money can go towards.

What Murray is sympathetic to is doing more for gifted students, because he thinks they are underesourced to realize their true potential. That isn't an argument for the average classroom needing gold standard schooling, because Murray thinks it ain't doing much.

[_honorentheos]

Other that because we like metrics, why should one privilege IQ as the measure of performance and how we ought to value education or anything at all? Just curious.

[_EAllusion]

Other that because we like metrics, why should one privilege IQ as the measure of performance and how we ought to value education or anything at all? Just curious.

The Bell Curve's argument is that IQ measures real underlying intelligence and high intelligence is a cause of desirable social outcomes and low intelligence is a cause of undesirable ones. No one disputes that IQ testing is reasonably predictive of those things, but the argument here is that IQ is a valid measurement of a underlying trait that substantially causes those things. Therefore, we should care a lot about it.

And it's not just that being good at mental tests is a helpful skill to have because it makes it easier for you to get into good colleges and whatnot. It's that being good at mental tests has a lot to do your natural ability independent of your social situation and this has important implications for your ability to be successful.