Genetic Russian Roulette

[TRIGGER WARNING: This would be a really bad post to read if you have or are about to have a young child]


One of the downsides to working in psychiatry is that it is slowly but inexorably sapping away first my enthusiasm about, and now even my willingness to, have children.

Medicine was bad enough. It wasn’t the kids who get leukaemia at age seven and die who got to me. It was the ones with syndromes. Down Syndrome is the one everyone knows about, but by no means the worst. Lesch-Nyhan Syndrome is a lovely little condition in which children are born without the ability to metabolize uric acid properly. The results include severe mental retardation, stereotyped jerking movements of the limbs, face frozen in a permanent grimace, and something the textbooks charmingly refer to as “involuntary writhing”. For poorly understood reasons, these children also exhibit “uncontrollable self-injury”, usually head-banging and trying to bite any part of their body within reach of their mouth – something caregivers quickly learn not to let parts of their bodies be. Oh, and also vomiting, spitting, and uncontrollable urges to use profanity.

(for a fun intellectual exercise, imagine what arguments might be able to convince William of Ockham that “deficiency in uric acid metabolism” is a more parsimonious explanation for this syndrome than “possessed by demons”)

The disability rights crowd will probably call me ableist, or disableophobic, or misodisablistic, or whatever people say these days, but the idea of having a child with Lesch-Nyhan Syndrome, and having to take care of him until he dies (likely in his teens; it’s not a very survivable disease) or gets old enough to go off to a group home – that terrifies me. I worry I would spend every second of every day hating my kid and begrudging him the countless hours and tens of thousands of dollars I would be spending on him, all while maintaining a smile both to him – because goodness knows he has it bad enough already without having to cope with crushing guilt and parental enmity – and to the rest of the world – so I could get my official status as Brave and Kind Long-Suffering Caretaker, instead of Ableist Jerk Who Kind Of Wishes He Could Just Put His Kid Out Of His Misery And Get On With His Life).

(note: I use the masculine pronoun not unreflectively, but because Lesch-Nyhan is x-linked recessive)

The good news is the incidence of Lesch-Nyhan Syndrome is only 1/400,000 births. The bad news is that if it’s not Lesch-Nyhan Syndrome, it could be Treacher-Collins Syndrome. Or DiGeorge Syndrome. Or Cornelia de Lange Syndrome. Or any of a thousand others. Have you ever seen a Pediatric Genetics text? They’re not small books.

And not all of these are Lesch-Nyhan-level bad – I like playing with Down Syndrome kids. But I would still be unexpectedly finding out that what I thought was just a perfectly normal super-gargantuan committment to sink most of my resources into a mostly-helpless creature for twenty-five years would in fact become, without my planning or consent, a super-ultra-titano-gargantuan commitment to sink even more of my resources into an even-more-helpless creature indefinitely. And all the associated dreams – of seeing my kid get grow up and raise a family, of debating the great questions of life with zir, of secretly training zir to be a child prodigy who does calculus at age five like John Stuart Mill – would all be gone. I am reluctant to say “all the things I look forward to about having a kid”, because someone will lecture me about how I should want to just gaze deep into my child’s eyes, loving them unconditionally as a human being in the image of God. But as the saying goes, we can do what we want but cannot want what we want. Also, eye contact is scary.

But at this point it’s not even the syndromes that worry me the most. The kids with syndromes I’ve met are nice people, I like them, and if it came to that I’d at least have great social support from my amazing friends.

Right now what scares me is psychiatry.


I remember vividly the first time I met my first antisocial personality disorder patient with perfect parents.

The dad was a something something manager at Ford. The mom was a high school teacher. Both from middle class backgrounds, never been in a fight, never been in jail, never even used drugs. Two other children at home, one of them working on college applications to nice schools, the other getting As in junior high and playing in the band.

Their oldest kid was in the psychiatric hospital where I worked because…actually, no, I can’t remember what he was in for, the particular time when I met his parents. It could have been the stabbing people. It could have been the constant drug use. It could have been lying like a rug to the police. And as I was explaining whichever the latest disaster was to his parents, on their faces was just this look of heartbreak mixed with total lack of surprise.

I’d seen dozens of patients worse than this guy already, but they’d all been easy to brush off as something that could never happen to me. Either their parents were just as bad as they are, or they’d been beaten as a kid, or they had a history of sexual abuse, or something else where it was very easy to say “Aha! I will just not attack my children with broken beer bottles when I am a father, and then they won’t turn out this way.” Or since technically I’m supposed to be kind of a genetic determinist, I could just not have terrible person genes, and marry someone else without terrible person genes, and our kid would be okay too.

This guy’s family – and he was the first of several – didn’t give me any of those easy outs. You can do everything right in the world, and your kid can still grow up to be a murderer, or a rapist, or just one of those guys with a tattoo of a skull and a nickname like “Snake”.

I like the metaphor of “the genetic lottery”, but on the worst days it starts to feel more like genetic Russian Roulette. Maybe the first kid is beautiful and grows up to be a scientist, the second one is compassionate and becomes an artist, and then boom, the third one vivisects stray cats and tries to burn the house down. But it’s too late to return her for store credit; you’re morally obligated to spend twenty years taking care of her and sending her to special schools and taking late night calls from the police and paying her bail, all in the desperate hope that maybe one day she’ll shape up. Like I said. Russian roulette.


I get irrationally angry whenever I hear people diss discipline. Like, if you just believe spanking is often bad for children, then that’s fine and as far as I can tell empirically correct. What bothers me is the people who see someone spank their kid and say “What that parent really needs to learn to do is talk things out. If they would just learn to communicate with their kid, the kid would learn that what ze’s doing is wrong, apologize, and none of this violence would be necessary.” This is the sort of statement that makes perfect sense if you’ve never been in a position where you had to control a genuinely bad (or even mediocre) kid.

(as such I am suspicious of its signaling value. That is, I worry that saying children should never be disciplined harshly is a way of saying “Ha ha, my children are so great I never need to discipline them. And there you are with a bad apple. Chump!”)

Don’t get me wrong, there are lots of problems with parents abusing kids under the excuse of “discipline”, and the idea of an unelected and unaccountable individual being able to hurt someone else whenever they want is terrifying. It might be that a Congressional bill to ban all discipline stronger than dirty looks would on net be a good thing. But when people try to push it by saying “If you just demonstrated your love and affection, you wouldn’t need any discipline” – that’s what gets my blood boiling.

I think this dates from my time as a schoolteacher. When I was a student, I hated all my teachers and thought that if they just ditched the constant repetition, the cutesy but vapid games, the police state attitude, then everyone would learn a lot more and school would finally live up to its potential as “not totally incompatible with learning, sometimes”.

And then I started teaching English, tried presenting the actually interesting things about the English language at a reasonable pace as if I were talking to real human beings. And it was a disaster. I would give this really brilliant and lucid presentation of a fascinating concept, and then ask a basic question about it, and even though I had just explained it, no one in the class would even have been listening to it. They’d be too busy chattering to one another in the corner. So finally out of desperation I was like “Who wants to do some kind of idiotic activity in which we all pick English words and color them in and then do a stupid dance about them??!” (I may not have used those exact words) and sure enough everyone wanted to and at the end some of them sort of vaguely remembered the vocabulary.

By the end of the school year I had realized that nothing was getting learned without threatening a test on it later, nothing was getting learned regardless unless it was rote memorization of a few especially boring points, and that I could usually force students to sit still long enough to learn it if and only if I bribed them with vapid games at regular intervals.

Yet pretty much every day I see people saying “Schools are evil fascist institutions that deliberately avoid teaching students for sinister reasons. If you just inspire a love of learning in them, they’ll be thrilled to finally have new vistas to explore and they’ll go above and beyond what you possibly expected.”

To which the only answer is no they frickin’ won’t. Yes, there will be two or three who do. Probably you were one of them, or your kid is one of them, and you think everything should be centered around those people. Fine. That’s what home schooling is for. But there will also be oh so many who ask “Will the grandeur and beauty of the fathomless universe be on the test?”. And when you say that the true test is whether they feel connected to the tradition of inquiry into the mysteries of Nature, they’ll roll their eyes and secretly play Pokemon on their Nintendo DS thinking you can’t see it if it’s held kind of under their desk.

I don’t think I used to be an optimist. I think I used to be a narcissist. I figured that when I was a teacher, everything would work out, my kids would be kind and attentive, my lessons would stick, and there would be no behavioral problems or if there are they would quiet down after I give them a friendly talk about why attention is important. I felt like the Universe owed it to me to have everything work out. I didn’t realize on a gut level that kids could just not cooperate.

The parents who pooh-pooh strict discipline have the same blind spot. It’s an easy blind spot to have if your own kids are wonderful. It’s hard to realize that you can just ask them to behave until you’re blue in the face, set strict limits, send them to bed without dessert, and the kid can just choose to not behave.

And we’ve come a long way from Lesch-Nyhan Syndrome here, but these all tie together into a complicated knot of worry. It’s a fundamental realization that I could have a kid I can’t fix. Not in the “I want her to be a businessperson but she wants to be a poet” sort of way, but in the “I want her to not vivisect cats for fun, and she wants to vivisect cats for fun” sort of way. Someone whom I lavish all the love in the world on, and give exactly the right dose of prenatal iodine, and whatever, and the universe just says “Nope, you’re stuck with this person. You should have thought of that before one of your sperm had a single nucleotide mutation in an out-of-the-way corner of the genome”.

I have a serious girlfriend, and I’m only two years younger than my father was when he had me, and all my friends are having kids, and I really want children, and I have no idea what I’m going to do.

246 thoughts on “Genetic Russian Roulette

  1. Pingback: One Brain, Many Selves: Demons, Tulpas & Neurons Gone Wild

  2. Becky

    I just wanted to say that this post isn’t so terribly scary to those with/wanting young kids. I have a toddler and a one-month-old. I regret reading Andrew Solomon’s Far From the Tree when I was pregnant with said one-month-old. There’s nothing you’ve written that comes anywhere close to approximating that book’s horrors.

    Erm, so don’t read that book if your girlfriend wants kids. And please don’t string your girlfriend along if she wants kids and you’re ambivalent, lest she end up like a friend of mine who left her husband at the age of 40 because he wouldn’t change his mind about having a kid. She went with a sperm donor and is a single mom to a lovely toddler, but I don’t think that’s anyone’s ideal scenario.

    Having kids is scary and unpredictable; so is life. You could sire a child with a horrible disability. You might also die from a freak aneurism in five minutes. Most people, however, have healthy and reasonably functional children and live to be at least 70 or so. I have anxiety problems and tend to dwell on the scary “what-ifs,” but I’m much happier when I plug my ears and sing “la la la” to myself about such things.

  3. Pingback: Linkspam, 11/29/13 Edition — Radish Reviews

  4. spandrell

    Being an utilitarian will obviously discourage you from having children. You don’t really need to be aware of disease odds to understand it. If you calculate how cute they are or the joy the provide against the sheer hassle, unpleasantness, uncertainty and the plain loss of freedom they represent, child rearing it is seldom a good investment, especially in the modern era when you can’t depend in their money and company in your old age.

    The only way of having children is by pure biological inertia (your wife nags you and you agree without really thinking about it), or because of a sense of duty towards your tribe, nation, humanity or whatever moves you.

    You LWers could go a step further and nag each other into having children, as that would increase the number of smart/rationalist kids on earth, which if you all respond to and actually have the kids, will make life much more pleasant. It would make life easier for conscientious kids at school for example.

    So you either believe you have a duty to multiply your kind, or you consume your single life in barren utilitarian calculations.

    1. David Gerard

      I’m not particularly utilitarian – I suspect consequentialism is correct, but in practice I do virtue ethics in hardware ‘cos I’m a human – but I had “reproducen” as a fairly explicit terminal value. So, one kid I’m bringing up (and trying for another) and sperm donation. I could feel the nagging from my genes quieten.

      I can’t find the reference quickly, but I recall J.G. Ballard saying he lost all fear of death once he had grandchildren – like his genes were saying “well done, phenotype, job done.”

    2. ozymandias

      Erm, most utilitarians care about maximum utility for everyone, not just for themselves. The latter is generally called “enlightened self-interest” or something along those lines. So utilitarians could altruistically have babies.

      This comment section, in general, seems to be full of people who not only don’t want kids but cannot comprehend the mindset in which one would desire kids. This is really weird because I know a lot of people (including myself) with an intense desire for children– people like David Gerard who have reproducing as a terminal value. Am I having a case of very unusual selection bias and is Intense Desire For All The Babies, like, a queer feminist thing? (That would be weird.) Or have people not talked to other people about their desire for kids enough to figure out that some people do, in fact, very much want children?

      (…why is your wife nagging you for kids if kids are terrible and no one wants them)

      1. Randy M

        Not all wives, even of rationalists, are utilitarian.

        Most of life, for most people, is impulse, whim, and fancy.

      2. spandrell

        Women are supposed to nag about having babies. That’s what women are made for. The genes of women who don’t nag for babies go extinct very fast.
        For men it requires a change of mindset, a shift from pleasure seeking. I’d rather join my mates and go in a trip to Thailand, but I can’t.

        Kid’s aren’t terrible, they can be quite great. But once they’re out it doesn’t matter that they’re a pain in the ass who take away all your leisure. They’re yours and any halfway decent human being will do anything it takes out of love and duty.

      3. houseboatonstyx

        This comment section, in general, seems to be full of people who not only don’t want kids but cannot comprehend the mindset in which one would desire kids.

        I, on the other side, was seeing it as full of people pushing the pro-kids side, with no sympathy even for Scott’s undecided position.

        I wonder which, if either, of us will go to the trouble of counting the posters, and the posts, on each side. Maybe the pro’s are making like one post each, and the anti’s are hanging on arguing.

  5. BenSix

    I would be surprised if rationalists are not far less likely to have children than most people. Not just because rich, well-educated people are less likely to have children anyone but because parenting is so much about gut instinct and thoughtlessness. Otherwise one might go mad when faced with the scale of one’s responsibility, and the scale of the nonsense that is supposed to aid one.

    1. houseboatonstyx

      But the largest incidence of the very worst things that can happen to a child, are the sort of extreme physical conditions mentioned in the OP, and there’s not much a parent or prospective parent can do to prevent those.

      Once you’re actually parenting, the incidence of such extreme damage to a child becomes much smaller (accident with freak effects etc), and again there’s not much you can do to prevent freak effects. So as a parent, it makes sense to discount those most extreme nightmares, so as to do a better job with the more likely situations. And parents’ instincts are usually pretty good at knowing the difference.

      It’s at the decision whether to BECOME a parent, where you can prevent the most extreme risk to yourself: by not becoming a parent. So there it is reasonable to look at the actual odds, and at how bad the worst cases are.

  6. nydwracu

    To which the only answer is no they frickin’ won’t. Yes, there will be two or three who do.

    Are there reliable ways to find those two or three in every school? Are there enough of them that programs could be developed to find them and put them in their own school, one that won’t have to optimize for babysitting people who just want to play Pokemon? If so, an obvious solution presents itself—or would, if not for the likelihood of such a thing being impossible for political reasons. (cf. the death of tracking)

    This seems to be the closest thing that currently exists. (Magnet programs in ordinary public high schools don’t cut it. I was in one. Not only was there the problem of having to deal with genpop, who, you know, it certainly seems like it would be better to keep the smart people and the people who are going to get in fights in the parking lot and get in their cars and run each other over until someone bashes in their windshield and drags them out through the glass shards and beats them into unconsciousness separate; there’s also the problem where it’s limited to either a county or a certain area within the county. Mine served a third of my county and had about 300 slots, which was at least 250 too many. TJHSST, on the other hand, serves five counties and two cities.)

    1. nydwracu

      …then again, it may actually be a *good* idea to make sure smart people have firsthand experience with genpop, given how many I’ve met who have no such experience and whose political positions I suspect would change significantly if they had such experience… it’d be possible in theory to design political systems that wouldn’t benefit from throwing them into those situations, but good luck getting those implemented ever.

  7. hf

    Ana Mardoll and company say “Content Note”, which is shorter and clearer.

    No idea what to do about the issue that concerns you.

  8. Douglas Knight

    I wonder if you are mixing diagnosis numbers with observations of patients who are not representative of the diagnosis. Autism spectrum diagnosis is 1%, but it’s a pretty wide spectrum. I find it very hard to believe that prototypical ASPD is really 2%.

    1. Scott Alexander Post author

      That seems a bit catastrophic. Most bacteria still respond to most antibiotics. There are certainly some multi-drug resistant strains, but they tend to hang around hospitals and only get a foothold in people who are already very sick. I’m not saying it won’t increase the death rate for hospitalized old people or give you a bad time if you get an infection from an indwelling catheter, but at the moment there’s no reason to think that kids will be dropping like flies or anything.

      Put it this way. Right now we treat bacteria resistant to everything else with vancomycin. Have you needed vancomycin yet in your life? No? Then the appearance of strains that are resistant to vancomycin (what most people mean when they talk about omniresistant superbugs) wouldn’t have affected your life so far.

      (also, I predict that around the time all our last-line antibiotics are becoming exhausted, we’ll have PPMOs or something like them up and running)

      1. Nestor

        That was my take on it, although it was bacteriophages I’d heard as an antibiotic replacement, but then I read this article on it

        after reading this one

        Frankly it seems like a lesser existential threat than all out nucular war, which didn’t stop my parents from having me in the 70s (What were they thinking!) but it’s all additive, isn’t it?

  9. tadrinth

    I concluded a while back that if genetically engineering of human embryos is available, it would be wildly irresponsible and immoral to not genetically engineer my kids. I then tentatively concluded that if possible, I should wait until we have reasonably good genetic engineering to have kids. I’m not really in a good position to have kids at the moment, but if I was, I would be terribly conflicted over this, because I’m very unsure if I can realistically wait that long.

  10. houseboatonstyx

    Perhaps the rational procedure for someone in Scott’s position would be:

    For spreading his own genes, donate sperm.
    For a child to raise, adopt; whether a cuddly baby (who has had at least some health screening) or an older child in need (who has had more years for health screening, and may be in need of a psychiatrist for a parent). The mental state of the older child would not be just luck of the Foster Care pool. The person who wants to adopt can find that out before choosing which child.

    If it’s important for the child one raises to have one’s own genes — then what is the Rationalist reason for that?

    1. Elizabeth

      Genes and memes interact. Some people feel that it is important that their offspring get both from them.

  11. Carinthium

    Question- what actual percentage of children are born with at least one disorder? Shouldn’t you keep that in midn?

  12. coffeespoons

    Just to add, my guess is that most people with mild/moderate depression/anxiety have sufficient good times to make life worth living. I can’t actually find any data though.

      1. houseboatonstyx

        It might depend on estimates of what? And, and how much objective evidence is available.

        Scott has got both good information and first-hand observation of the worst, or very bad, conditions that can happen to the elderly and to the severely handicapped children — who are hidden away from the rest of us. Thus his OP assessment of the risk and its severity is likely to be more accurate than ours, who go along rejecting thoughts of disaster as ‘just depression’.

        1. Nestor

          Scott is king of figures. My own insights are more qualitative.

          For example, when I’m depressed people look like they’re made of raw meat.

          When I feel better, they don’t look like raw meat anymore. But I’m aware that in fact they are made of raw meat. The simple intellectual fact doesn’t hold the same weight as the raw emotional perception. But it’s not innacurate.

  13. coffeespoons

    I think Scott was right not to include depression/anxiety disorders. I have elements of anxiety, OCD and depression, and I think that me existing is a good thing, and that my life is very much worth living! I also think my parents got a lot out of raising me. However, I suppose depression that causes one to commit suicide/be constantly suicidal is the kind of thing that you wouldn’t want in a child. I’ve never been suicidal.

    1. Said Achmiz

      Severe depression includes, by diagnostic definition*, thinking that your life is not worth living. I would definitely not wish it on any child I’d have.

      However, it is not necessarily the case that being severely depressed for part of your life necessarily means that your life as a whole is not worth living. I can attest to the opposite.

      *Scott / other professional psychiatrists/psychologists, correct me if I’m wrong; I’ve only been on the other side of the table and may be misremembering.

      1. coffeespoons

        Nestor said “You forgot depression”, so I assumed he meant mild and moderate depression as well as severe depression.

        1. coffeespoons

          And yes, I agree with this:

          However, it is not necessarily the case that being severely depressed for part of your life necessarily means that your life as a whole is not worth living.

    1. Said Achmiz

      I don’t see any mention in the linked article about the victim having been raped. Is that mentioned elsewhere?

      1. Nestor

        Yes, I originally read a different article. The rapist tried to kill him to prevent him from identifying him.

        It’s just one example of utterly horrifying shit that can happen to a normal kid. Low probability of course, but then so are the syndromes talked about in the OP. Worse, in a way because your bright eyed and bushy tailed perfectly formed kid can get mangled by life in an instant, arguably the ones with crippling syndromes may not have enough self awareness to know the difference.

        1. houseboatonstyx

          your bright eyed and bushy tailed perfectly formed kid can get mangled by life in an instant

          Sadly. So the incidence of healthy children becoming helpless by later accident or illness, should be added to the base chance of birth defects or early-detected abnormalities.

  14. Krisis

    I’d say that creating a being that suffers immensely, like the L-N syndrome kid, would make me feel guilty, not just because of being unable to love zir, but because this grim suffering was caused by me.
    On the other hand, when you breed, you ALWAYS consciously create a sentient being that will suffer – that is inevitable – and then experiences the humiliation of decline and the terror of death.

    Having children is morally problematic even without syndromes or other genetic accidents. And furthermore, it’s really hard and unrewarding to be a good parent. It’s easy to rationalize that into “oh it’s so worth it” but it really isn’t. You’ve just created suffering and death, and it doesn’t make you feel good about yourself. Don’t breed, try to live a good life instead.

    1. Said Achmiz

      Are you claiming that life is, on net, not worth living? If so, do you say that it’s:

      a)All people’s lives are not worth living on net?
      b) The average person’s?
      c) The average person’s in some relevant reference class that includes Scott’s likely offspring?

      I would probably have to disagree with a), agree with b), and disagree with c). To take a personal example, thus far, I would say that my life has been worth living.

      This was not always the case (which is to say, if you were to terminate my life at various points e.g. 5, 10, 15 years in the past, it would not always have been the case that my life as a whole will have been worth living), but it is true now, and I expect it to become progressively more true going forward.

      I expect that Scott would likely be able to ensure, with high probability, that any children he has live lives that are, on net, worth living, both for the majority of their expected lifespans and taken as a whole over the course of their lives.

      1. Krisis

        I’m not claiming anything like that. I think we are hardwired to make sense of our lives and to invent meaning even of the most grim circumstances. Even if Scott would have a kid that has some kind of issue (not unlikely it seems, was it 13% chance?), Scott would adapt and attach meaning to all the troubles and not regret a thing. I just think that childlessness would end up the same way, rationalized into something great and meaningful. So, abstaining from the roulette is the better choice.

        (Now, in the deeper sense I think it is problematic to create someone who will suffer, decay and die (that is, any human being) – even if that person is well-functioning enough to experience zir life as worthwhile, because deep down we all understand that it is horrifyingly pointless, but our survival instincts will not let us entertain that unbearable simple fact. But that’s just my opinion.)

  15. Anonymous

    This was a great post.
    A good opportunity to thank you for the brilliant stuff you write.
    Thank you.

  16. Avantika

    I have no great advice to offer on the child problem beyond sympathy, but thank you for totally summarizing all the reasons why I no longer want to be a teacher.

    1. Scott Alexander Post author

      I support disability rights, I support active disability rights organizations like the National Alliance on Mental Illness and the Epilepsy Society, and I try hard to stand up for people with disabilities at my job (which is not always a costless or popular action)

      My feelings are much less positive towards some of the people who self-identify as “the disability rights movement” on social media or in the blogosphere. I am trying to be less angry all the time so I would prefer not to expand on this further.

      Also, was that asked in a confrontational tone, or am I just imagining it? Hard to read this things over the Internet.

      1. Patrick Robotham

        To be honest, yes.

        I would like to confront you further on this point, but feel like I can’t decently do so in light of your general loveliness and stated preferences.

      2. Patrick Robotham

        I’m terribly sorry Scott. I felt guilty for attacking your liberal credentials and your reply reminded me that this is unfair and that you’re a lovely person. I also felt afraid that you’d be angry at me if I confronted you. This guilt and fear got turned into a passive-aggressive and condescending insinuation that you’re overly sensitive and unable to have a proper discussion about social justice.

        I shall now use a cleansing ritual: Scott, you’re a very lovely person and your liberal credentials are impeccable.

        On to the confrontation!

        While I wouldn’t dream of accusing you of failing to be either a) liberal, or b) lovely, I do think you’re doing two bad things when you’re being snarky at social justice warriors (in this post, social justice warriors subspecies ‘disability rights movement’) that you find lurking in the blogosphere. You’re being uncharitable (sin of sins!) and you’re helping those who reject the concept of social justice itself (get over it Patrick!)

        I think not being charitable is a big deal, and most of this comment is dedicated to why I think you’re being uncharitable.

        The other thing I mention only because it’s affected me personally, and was probably the reason for my earlier confrontational tone.

        1. Being uncharitable:

        As unpleasant and inarticulate as these social media types can be, their feelings, attitudes and positions can be rooted in cogent arguments.

        Take for example, your complaint in
        “admiring an unattainable woman from afar and showering her with presents as an expression of their transcendent yet hopeless love. Or, as we moderns call it, being a Nice Guy (TM) and therefore Worse Than Hitler (TM).”

        The Worse Than Hitler (TM) is charmingly dismissive hyperbole, and I think an internet meme. But in your haste to be snarky, you got the definition of “Nice Guy” wrong. There’s an interesting discussion to had about precisely how bad are the people who think “courtesy entitles me to sex”. They’re clearly not worse than Hitler, but they might be worse than muscular people who wear fake tan. But you shut down that discussion with your tradesnark.

        I don’t mean to dredge up the past in order to yell at you about feminism. You’ve acknowledged that you misunderstood the term after all. What I worry about is you doing the same thing to “Ableist” that you did to “Nice Guy”.

        “People who think that being a parent of a disabled child is sucky are Worse Than Hitler (TM)” and “People who think that nobody would want to be disabled are Worse Than Hitler (TM)” might both strike you as status-seeking, bully boy positions. But they are different positions. You’ve thoroughly discredited the first. You can defend a version of the second (once you stop falling afoul of Godwin’s law), and I’d wager the blogosphere “disability rights movement” advocate the second.

        Scott, you’ve passionately advocated charity in argument. It’s one of the reasons you’re such a lovely person. I wish to thank you personally for this advocacy. I’ve taken it to heart (even if I don’t always practice it). I now occasionally have arguments inside my head about some controversial issue and I don’t know who the winner will be. Furthermore, my boyfriend informs me that being charitable in my arguments is one of my virtues and a reason to love me.

        Scott, I still think you’re wonderful,
        <3 Patrick Robotham.

        1. Scott Alexander Post author

          Give me an email address (you can send to scott [at] shireroth [dot] org) and I’ll discuss this with you in private.

  17. JRM

    1. This was a great post. Just awesomely great.

    2. roystgnr’s comments about your intuition being off is right. You’re not seeing the denominator. The denominator problem is one I seldom see addressed, but it’s here. Yes, there are kids who are trouble – big trouble – with good parents. (I’m a prosecutor. Yeah, it happens.) But percentage-wise? You’re in a way-plus position.

    3. If you want kids, have kids. (Get married first. Then have kids. That’s really the right order.) And do the genetic testing in utero, and if they have a serious terrible condition, abort. If they end up as a psych case…. well, life is risk. But this is a very low risk.

    4. I had a nephew with Asperger’s commit suicide. I get that there are risks intellectually and viscerally. But you have to look at the denominator of the kids that are generated by people similar to you. The demographics of those kids is really, really good overall.

    Good luck.


  18. Grognor

    The correct solution to this problem is antinatalism. Only one person pointed this out, but he did in the worst possible way so I thought I’d do it better this way.

    I know you know antinatalism is a thing. Why didn’t you even mention it in this post?

    1. Said Achmiz

      It seems as if Scott experiences (as do many people, I gather) a desire to have children. I do not understand it, myself, and agree that antinatalism is the correct solution; nonetheless, we have to acknowledge that actively wishing to procreate (as distinct from simply doing because, well, that’s what you do) is apparently a pretty strong drive in a sizeable portion of the populace.

    2. Scott Alexander Post author

      These worries seem orthogonal to antinatalism. Antinatalism as I understand it is a position that says it’s morally wrong to have children. I reject that position, but I am uncertain on the different question of whether it’s dangerous/unpleasant to have children.

  19. Noah

    You should not have children. You are not doing them a favor, and you are not doing yourself a favor, either.

    Spend the money on yourself and enjoy your life instead!

    (If you want children because you are a pedophile, you should remember what they do to people like you in prison.)

  20. Emile

    Whew, I’m relieved there’s nothing too scary here, the trigger warning made me expect worse (I have a three year old).

    It’s a fundamental realization that I could have a kid I can’t fix.

    Sure, and you can have problems you can’t solve, you can catch diseases you can’t cure, you can run into obstacles you can’t overcome.

    BUT, you’re also likely to have a kid you can fix better than most parents could. How many parents do you think could tell better bedtime stories than you? Also, think of all the Legos you could build! (Did you know you can make a pretty wide variety of mechas with standard Duplo blocks?) The weird things to teach!

    Maybe I should post anecdotes about my kid here all the time. The other day we were coming back from the park, he was on his little push-bike (no pedals). We pass a couple of cute teenage girls chatting on a park. He stops in front of them and proudly announces “this is my motorcycle!”.

  21. Ben

    You’re a doctor. If you have a kid whose life isn’t worth living, or makes your life not worth living, surely you could arrange an unfortunate interaction between medications halfway through a transAtlantic cruise.

      1. Noah

        But thank god it’s okay to shit them into an unchosen existence so that they can suffer for decades without any shred of informed consent!

      2. Said Achmiz

        Yeah, I’m not entirely sure I see what’s wrong with this proposal. Is it the whole “right to self-determination” thing? If so, I assume Ben was referring to those cases where the child in question does not have (and will never have) the cognitive capacity for such self-determination.

        What is the argument against killing such a person? (An interesting parallel to consider might be to cases where the relatives of a person in a coma or with brain damage etc. have to decide whether to turn off life support.)

        Naturally, in cases where the kid is disabled in some way but is still able to express a desire to continue living, killing them would have the same moral implications as killing anyone else. Other cases where the option would be off the table would be, for example, non-verbal but cognitively within-human-range autistic people.

        P.S. Noah’s point, while not (imo) a knockdown argument, must unquestionably be addressed in any account of the moral status of the severely disabled.

      3. ozymandias

        Ben said “a kid who makes your life not worth living.” He did not say “a child who does not want to be alive and consents to euthanasia and is capable of making this decision freely” nor did he say “a child living a life that, in your best assessment, is not worth living and who is not cognitively capable of giving consent.” He didn’t even just say “a child whose life is not worth living.” He literally suggested murdering a *person* because they make your life suck.

        It is not okay to murder people! Especially children! What the fuck is wrong with you?

        I *do* support seeking to non-coercively eliminate genetic diseases, but I do not actually think this is contradictory to a strong anti-child-murder position.

        1. Noah

          Surely the problem could be solved by

          1) legalizing the killing in the cases you described (consent given, permanently incapable of giving consent and life probably not worth living)

          and 2) absolving people from obligations to others who can’t care for themselves.

          Basically the parent could say, “I no longer wish to be responsible for my child”, and then other people could volunteer to take care of them and cover the expenses. And if no one volunteers, let them die. Omitting permanent care is not murder. Otherwise we’d all be moral slaves to those who can’t live on their own.

        2. Said Achmiz

          Ben said […]

          We practice the principle of charity here, no? I think we can infer Ben’s roughly intended meaning. (Of course, he could also chime in and clarify.)

          It is not okay to murder people!


          I don’t ask just to be contrarian. I think murder is wrong in general. But it’s wrong for specific reasons, not, like, just because. And sometimes those reasons don’t apply. And then it’s not wrong anymore.

          Especially children!

          Now here I totally disagree, with no irony. I don’t think there’s anything specially bad about murdering children. You shouldn’t murder people, in general. But children are not somehow more worthwhile than adults. I actually find the view that they are… distressing and insulting, to put it tactfully.

        3. Said Achmiz


          and 2) absolving people from obligations to others who can’t care for themselves.

          Hmm. I don’t think that I’d prefer to live in a world where someone who had a responsibility to care for someone could abdicate that responsibility at will. I think the relevant questions that would determine the viability of such a thing would be:

          What institutions exist to take care of people who can’t do for themselves? What are they like (more similar to expensive private hospices or more similar to Child Protective Services and such organizations)? Does the abdication-at-will rule apply to them?

          And probably others I am failing to think of at the moment.

        4. Noah

          Said, I think it makes sense to hold people who consented to take care of other people responsible for their promise. Especially professionals who get paid for that.

          Other than that, I’m against such a forced responsibility. For instance, I don’t think anyone should be taxed for it. There are always going to be people like ozymandias who consider it their honor to take on the responsibility voluntarily.

          The case of parenthood is somewhat special. I guess it could be framed as consenting to caretaker responsibility by having sex, especially unprotected sex, or donating sperm, knowing there’s at least a risk of disability.

        5. ozymandias

          But… our society *does* provide permanent care to those who can’t take care of themselves. It’s called Social Security disability. That’s not an “unless” that is a real thing that currently happens. I wouldn’t call that being a moral slave or, indeed, particularly harmful to me at all. (The taxes are more than outweighed by the reassurance that if I get hit by a car and get paralyzed from the waist down I won’t starve to death, as well as that my disabled friends will be taken care of.)

          You also have the right to ‘get rid’ of your child without murder if you so desire. You can voluntarily place your child in foster care. While foster care is suboptimal, it’s generally better than being dead.

          Killing children generally involves the loss of more life-years than killing non-children (although I admit this doesn’t apply in the case of severely disabled children). In addition, I believe it makes sense to strongly condemn people hurting the weak and vulnerable, as well as those who are dependent on you, because they are less capable of defending themselves and therefore require more social stigma against hurting them.

          There is a strong Chesterton’s fence against killing people without their consent.

          1. gwern

            If SSDI really covers all that, why do we hear of so few people with the worst genetic disabilities being left to SSDI by their families? Maybe it doesn’t cover as much as you think it covers?

        6. Gilbert

          Ozy, your conclusion is correct, but there is no chance of squaring it with the philosophy you share with your interlocutors.

          If being a person is a question of intelligence, the folks here are just being logical, because we routinely kill animals more intelligent than kids suffering from the genetic disabilities in question. Granting that premise you don’t have a shred of a case.

          Now if you think killing homo sapiens is wrong even if they are dumber than animals, you are obviously right but let me be very blunt:
          That particular fence was burned down in 1973, you are clearly against rebuilding it and putting up a new one in the place you favor would be arbitrary and hypocritical.

        7. Earnest_Peer

          @Gilbert: Seeing how Ozy is vegan (correct?), I don’t think ey completely shares said philosophical underpinnings.

        8. ozymandias

          I am vegan but I’m pretty awful at it tbh (sushi is delicious). I also don’t think infanticide is necessarily immoral. Birth seems like a pretty sensible Schelling point about when not to kill humans, comfortably before they could reasonably be conscious; I would like to err on the side of not killing conscious beings. However, if another society has a different Schelling point I would not necessarily object. So I disagree with Gilbert on several points, although I am pleased to discover that *someone* agrees with me re: not murdering disabled two-year-olds.

          I also know several people who have lives that could be described as “not worth living” who think their lives are quite worth living actually, so I am somewhat suspicious of people’s definitions of “life not worth living.”

    1. Scott Alexander Post author

      It’s unethical for doctors to handle medication prescriptions for their own children.

      (and also to murder kids, I’m pretty sure)

      1. Said Achmiz

        Seems to me like that ethical restriction is in place to prevent harm to the child, no? If you’ve already decided to harm said child, then the ethical restriction against handling their medication prescription is rendered moot.

        1. Randy M

          That’s like saying “The laws against speeding are in place to avoid dangerous accidents. If one has decided to sideswipe a few pedestrians on the way home, the laws against speeding are moot.”
          Perhaps, but that doesn’t make us less sure of the latter than the former.

        2. Said Achmiz

          Um, I didn’t say anything about how sure we are of either thing. (Perhaps you thought my comment was intended as a reply to yours? I was actually replying directly to Scott.)

          My point was simply that the restriction against doctors prescribing meds to their own kids, while quite sensible as a general heuristic, does not apply here because our concern that it violating it might cause harm is overridden by our knowledge that our actions will in fact cause harm (and indeed that is the whole point of said actions).

        3. Said Achmiz

          Addendum: in other words, our motivation for having the rule in the first place (the desire to avoid harm) is absent in this case. We want to cause harm. It’s clearly silly to say “but your actions violate this rule we have that exists to prevent harm!”. Well, duh.

        4. Randy M

          “Perhaps you thought my comment was intended as a reply to yours?”
          I did, thanks for clarifying. Comment threading only goes so far.

      2. Doug S.

        Huh. My mom refills my antidepressant prescription all the time when I run out and my psychiatrist can’t be contacted immediately.

  22. Sarah

    I’m planning to have kids, and I’m *not* worried.

    There are a ton of genetic disorders but many of them are rare. Even autism, one of the commoner developmental disorders, is only 1% of the population (and there are quite a few autistics who grow up to do OK.) I’m not carrying anything that I know of, neither is my boyfriend, and we’re planning to get a more thorough genetic test after getting married and before we start trying for a baby [assuming all goes well until then.] Personality disorders that show up in childhood just don’t seem common enough to worry about. Keep in mind that you are a psychiatrist and your job is working with people who have psychiatric problems.

    Then there’s just happening to have an annoying kid I don’t like much. Or a kid less bright than my ideal. That’s common. I seriously doubt it would keep me from loving the kid. You get all those hormones for a reason.

    My current attitude is that affluent, elite-educated Americans are too obsessed with perfecting their kids. A lot of the stress and the expense is *optional*. Not actually necessary for raising a happy, healthy, decent human being who can live on their own; not even *really* necessary or sufficient for keeping the kid in the top social class bracket. Just a propitiatory measure against the apparent chaos in the world. I know I have the impulse myself, but I’ve also learned to be skeptical of it. And it really seems that raising children is less stressful and expensive if a.) you don’t try to do too much yourself, and b.) you don’t worry so much about optimizing the kid’s environment or behavior, beyond basics like safety and enforcing non-violent behavior.

  23. Anonymous

    But it’s too late to return her for store credit; you’re morally obligated to spend twenty years taking care of her and sending her to special schools and taking late night calls from the police and paying her bail
    No, you’re not morally obligated. You’re legally obligated. Just like they are legally obligated to go to school, to live with you, to keep living and suffering in the first place. Not morally, just legally.

    In a free society, I’d say go for it. Of course, in reality, there is no such thing.

  24. Rolf Andreassen

    Maybe you’ve done this already, but perhaps it would help to take a step back from what you see on a daily basis – I imagine you’re not exactly getting a representative sample of kids – and try to figure out the true odds of getting a bad ‘un? And, before you do, have a good think about what odds you’d be prepared to accept. Say, something like “If the chance of an unfixable kid with problems of size X is greater than 0.1%, I’m outta here.” It’s probably important to set the threshold before you do the math.

    1. Scott Alexander Post author

      See response to roystgnr above – odds of potentially bad stuff look somewhere between 5% and 13%.

      1. Rolf Andreassen

        I admit that this is a somewhat daunting number, if accurate. It seems to me that you did not adjust the Down’s incidence for the age of the potential mother? Apparently Down’s births are pretty heavily weighted towards older mothers, so that 1-in-700 average may be misleading for your particular case. (So I did a bit of research on this when my wife was pregnant and had similar worries…) Then, as someone pointed out, you should also adjust for the possibility of abortion. Similar comments may apply to the other syndromes you mention.

  25. lmm

    Is it the objective failure you’re worried about, or the subjective hating your kid?

    Anything you might possibly do has a certain chance of turning out to be pointless. I guess putting a lot of effort into a single project means the failure scenario is quite concentrated. But I’d imagine the odds that e.g. your entire research area is pointless would be comparable.

    As for the hating, did those parents you saw with the one bad son seem to hate him? What’s the outside view on your odds?

    1. Scott Alexander Post author

      They did an okay job of not looking like they hated him in front of their kid’s psychiatrist. I’m not sure how many conclusions to draw from that.

  26. Mark

    Given that you sound like you want to go on the journey, why are you thinking about driving there instead of taking the plane? You work for the plane crash cleanup company so you’re getting a bad view of the likelihood of dying in a plane crash vs. dying in a car crash.

    But even though the statistics may be sound, I totally understand your fear. Especially with my first child, it was very scary and a huge relief when nothing appeared to be wrong. Having an unhealthy child will be a huge fear up until they’re born and all the tests have been done. I have one friend who lost an infant (I’m not sure what the condition was) and another with a severely autistic child. So the genetic roulette probably isn’t a bad analogy.

    For me though, doing the expected utility calculation said have the kids and I bet it does for you too. I’ve made pretty good decisions in my life overall and really like where I’m at but I still sometimes question whether I went to the right school, choose the right career, etc. The choice I question the least though is choosing to have my kids. (Keeping in mind, this is from someone who appears to have not lost in roulette).

    1. Elizabeth

      > Given that you sound like you want to go on the journey, why are you thinking about driving there instead of taking the plane? You work for the plane crash cleanup company so you’re getting a bad view of the likelihood of dying in a plane crash vs. dying in a car crash.
      I don’t understand what the planes and cars stand for.

        1. Elizabeth

          You answered a different question from the one I asked. What do the planes and cars stand for? That is, if you weren’t using a metaphor, what would you say in place of those words? Planes seem maybe analogous to reproducing but I have no idea what cars are analogous to.

  27. Apprentice

    And then I started teaching English, tried presenting the actually interesting things about the English language at a reasonable pace as if I were talking to real human beings. And it was a disaster.

    Yeah, me too, though in my case it was physics. One of my clever ideas to make everything better was to put fun and challenging bonus questions on the tests. Something which the kids would be able to solve with a little creativity and without relying on rote memorization. This was my Waterloo. The smart-but-lazy kids benefitted so sometimes they would get a better outcome than the hard-working kids. You know, the hard-working kids – those precious few who reliably showed up in class, payed attention and did their homework. Alienating them was really, really not the way to improve the class.

    So, yeah, school is boring and stupid for a reason.

    1. Said Achmiz

      Why is it bad that the smart-but-lazy kids do better? In my experience, smart-but-lazy people do better in the truly high-end intellectual endeavors in real life, too.

      Speaking as a smart-but-lazy person, I can say that the sort of class/test design you describe is one of those times when I not only didn’t hate school, but actually enjoyed it. (Even better, of course, were those times when I could apply my intelligence/creativity the entire time, not just on bonus things, and didn’t have to do stupid rote memorization. Those classes remain fond memories, and sources of skills and knowledge that I still apply regularly in my professional career*.)

      *Like the 7th grade “math” class that we spent alternately playing with computers and building polyhedra out of construction paper. The resulting affinity and skills in technology, trigonometry, and spatial reasoning has served me VERY well indeed.

      1. Eric Rall

        It’s a question of incentives and modifiable behavior. The teacher wants to get the students to learn as much as they reasonably can. If you set up an incentive structure to try to motivate traits in your students towards this end, you’ll likely do better by incentivizing diligence than incentivizing raw intelligence.

        1. Said Achmiz

          Er, you can’t “motivate” diligence, smartness, laziness, or anything else in that vein. You can incentivize them, but that just ends up rewarding some people and punishing others.

          You can set up an incentive structure such that some subset of your students learn well; for instance, you can incentivize the smart-but-lazy ones to learn well, or the diligent ones; but you can’t transform one type of kid into another.

          As someone who tends toward smart-but-lazy, I submit to you that it would be Better™ to incentivize learning in people like me, in those subjects where raw intelligence pays off well, and where smart-but-lazy is the way toward going into that field and succeeding in it.

        2. Eric Rall

          I also tend towards smart-but-lazy, but I disagree with your premise that levels of diligent behavior are completely unresponsive to incentives. For example, I know from my own experience in school that I would be much more likely to complete an assignment on time if I faced steeper grade penalties for turning it in late.

          However, both the people firmly categorizable in the “smart-but-lazy” and “highly diligent-by-temperment” buckets aren’t the ones who matter when designing an incentive structure. Both buckets are likely to master the material in any reasonable classroom environment. It’s the people whose combined levels of dillgent-temperment and intelligence who are most likely to master the material only if the instructor provides an incentive structure that induces them to push themselves a bit harder than they naturally would by temperment.

  28. Apprentice

    I worry I would spend every second of every day hating my kid and begrudging him the countless hours and tens of thousands of dollars I would be spending on him

    This used to worry me too. What if an elitist nazi asshole like me has a child and it turns out to be disabled? Wouldn’t I hate the child? Well, then we did have a child and she did turn up autistic and with an IQ well south of 70. But I love her more than anything and everyone tells me I’m a great dad. Even though I’m still the same reactionary asshole. I bet Scott would be even better at this.

    I used to think I could deal with a physically disabled child but a mentally disabled one would be a bridge too far. How could we have any sort of connection? But it turns out that I have a lot in common with my autistic four-year-old. We enjoy some of the same computer games. I laugh honestly at her jokes and little pranks. Offspring are pretty fun, even if they turn out very different from what you’d imagined.

    (Okay, you could probably still have a full-on worst-case scenario with a really, truly impossible-to-like child. But I now think this is much less likely than I once imagined.)

    1. Scott Alexander Post author

      Thank you. This is one of the things I was hoping to hear, and it helps. There is a small but nonzero chance this comment will be responsible for me eventually having children.

      1. Apprentice

        If you are so inclined, I’d be happy to discuss this in more detail privately – including how the “disability rights” people are actively working to make our lives worse.

  29. jaimeastorga2000

    To which the only answer is no they frickin’ won’t. Yes, there will be two or three who do. Probably you were one of them, or your kid is one of them, and you think everything should be centered around those people. Fine. That’s what home schooling is for.

    I disagree. I see no value in sending these kids to a building for 6 hours per day, 5 days a week, year after year just so they can half-remember some useless trivia they memorized by rote a decade later. If school is to have any value at all as a center for learning math and science and language, it should be geared towards people like us and it should be expected that most students don’t belong there. Otherwise, if we are going to require that everyone goes to it, just dispense with the pretense of university preparation and turn it into a trade school system so the kids can practice woodworking, mechanics, cooking, sewing, knitting, etc… as a core curriculum. Or maybe use a branching system with diverging academic and practical education paths; IQ testing would do a pretty good job of sorting kids into the right path pretty early on, with some switch-over allowed for late bloomers.

    Not in the “I want her to be a businessperson but she wants to be a poet” sort of way, but in the “I want her to not vivisect cats for fun, and she wants to vivisect cats for fun” sort of way.

    Isn’t this what disownment and disinheritance are for? That’s why you should always have an heir and a spare.

    1. Said Achmiz

      I was planning to post almost exactly this, so I’ll just add my agreement and a corollary question:

      How effective is school / education for those kids who don’t want to learn, and are only “tricked” into “learning” by “vapid games” etc.?

      Because if it turns out that they don’t want to learn (and so don’t respond to techniques aimed at kids who do want to learn), and they aren’t even benefited by our attempts to force some learning into them, then we have a really solid argument for not forcing them to be in the education system.

    2. ozymandias

      You aren’t allowed to disown people under 18, so that plan would still leave you with nearly two decades of supporting and trying to take care of Cat Murderer Lady. You can put them up for adoption, but a twelve-year-old who wants to dissect cats is not going to be easily adoptable, even if they’re white, so you’re going to be leaving them to the kindness of the foster care system.

      Also, parents usually love their children, even their cat-murdery children, and do not want to get rid of them. I don’t have a child, but I noticed in myself a twinge of horror at the idea of *getting rid* of a child. This impulse seems supported by both parental anecdote and evolutionary logic.

      1. MugaSofer

        >I noticed in myself a twinge of horror at the idea of *getting rid* of a child.

        Thanks for having already said this so I didn’t have to.

    3. Anonymous

      While most people absorb much, much less from the school system than we might hope, I think it’s important still not to underestimate how much, well, dumber the average person who *doesn’t* receive an attempt at decent education tends to turn out.

      Our current educational system probably isn’t really optimized for, well, anything. It’s a hack job pieced together without clear goals or strong feedback mechanisms. But it’s Still a far cry from actual uselessness.

      1. jaimeastorga2000

        While most people absorb much, much less from the school system than we might hope, I think it’s important still not to underestimate how much, well, dumber the average person who *doesn’t* receive an attempt at decent education tends to turn out.

        Oh, definitely; all other things being equal, a high school graduate is probably better human capital than someone who never attended high school, or was home schooled, or became an autodidact. The important question is, do you think that’s more likely to be because of positive changes such an attempt at a decent education makes upon the person, or because of already existing characteristics which differ in those who are likely to receive such an attempt and those who don’t?

        1. Randy M

          “all other things being equal, a high school graduate is probably better human capital than someone who never attended high school, or was home schooled”

          Um, really? Don’t homeschooled as a group way outperform public school children? Yes, selection bias, but still, it doesn’t seem remotely a hinderance, nor do a see how it plausibly would.

        2. ozymandias

          The studies that suggest that generally have fairly poor methodology IIRC. So basically we don’t know whether homeschooled kids do better than non-homeschooled.

        3. jaimeastorga2000

          @Randy M: I meant that someone who had never attended high school AND was never homeschooled by his parents/tutors AND was not an autodidact in the style of Yudkowsky was probably worse human capital than someone who had done one of these things. The question is whether these things change the person for the better or whether already capable people are channeled into one of the three.

        4. Anonymous

          I think there’s more to it than selection effects, yes. I think the gap between the average person who went to a decent high school and the average person who went to a *non-decent* one is greater than just the gap caused by differences in non-school environment between the two.

          The book A Hope in the Unseen, a memoir which doubles as a look at the Affirmative Action system functioning as-intended, follows an individual who proves by the end of the book to be quite capable of doing well by the expectations of an Ivy League university, but only after a truly frantic, breakneck paced game up catchup to get to the level of the other students, because his own education so far hadn’t prepared him for demands of that level. This was a guy who was both bright, and at least a few deviations above the norm in terms of conscientiousness, committed to success even if it came at the cost of having a social life at his own school, who was able to perform at Ivy League standard after a few months, but only after a few months *at* an Ivy League school, despite his efforts at preparation.

          There are other pieces of data which suggest to me that school environment makes a difference for average students, of course, that piece alone certainly doesn’t clinch it, but not all my pieces of experience sum up easily in a blog comment. Volunteering at a nonprofit teaching kids in an underprivileged school district, and getting a chance to see a lot of their families, helps here.

          If you’re ever tempted to view the public education system *entirely* as a filtering mechanism, and think that the differences between the public in countries with decent ones and countries with crappy ones (remember Scott’s posts on Haiti?) are entirely down to non-school environment and/or gene pool differences, remember that for most of history, the vast majority of everyone has been illiterate. A Catholic saint was once canonized with one of his miracles being the achievement of learning to read before twelve. It was once thought of as Difficult Stuff for Smart People To Learn, now even dumbasses can do it.

    4. Scott Alexander Post author

      I don’t really disagree with you. School as it currently exists is probably a local maximum – the best it can do given political constraints that force it to be the way it is (eg every parent demands their kid be there during most of the workday for free childcare, people will get angry if children are stratified based on ability).

      There are ways of improving it by disrupting the very concept of school. You just can’t keep the setup the same and except a different teaching technique here or there to do much.

      1. Viliam Búr

        There is an option: separate children based on their ability in a way that does not rub their (parents’) noses in it.

        Create schools that look exactly the same from outside. Inside, provide one kind of education to the smart kids, and other kind of education to the stupid and lazy kids. At the end, give each of them the same diploma and pretend there was never a difference. This way everyone will get what they want — the smart ones will get the education, and the stupid ones will get the satisfaction that they were not separated away from the smart ones.

        You can have classrooms called A and B, you are just not allowed to openly speak about the fact that A is an elite classroom, and B is where you put the rest. Actually, you can split children differently at each subject, which will be good for students interested in one subject and ignoring another one, and will reinforce the illusion that the children are actually not separated.

        (I have heard a rumor that this is a part of the success of school system in Finland, but I have no data to support it.)

        1. Kaj Sotala

          High school math works roughly like that in Finland. The students can choose between taking “short” or “long” math, where the short version involves a smaller number of compulsory courses which also have simpler content. Some foreign language options are also offered in multiple versions. It is very obvious ability stratification and everyone knows it, but AFAIK nobody cares because it’s based on the students’ own choice rather than something externally imposed. (And yes, the “fundamental dignity of the human person” crowd is strong in Finland too.) You can also switch between the two versions later on, though switching from long to short is much more common than vice versa.

          The same is actually done even more subtly with the science subjects. The way Finnish high school works is that you have to complete a certain number of courses in order to graduate. Some courses are obligatory for everyone, for the rest you’re free to choose from whatever your school happens to offer. At least when I was in high school, physics and chemistry both only had one compulsory-for-everyone course, but lots of optional courses. (I think we had something like 10 physics courses available in total.) So if you weren’t good at science, you could just fill your high school diploma with courses you felt were easier.

          Everyone gets basically the same diploma at the end, but IIRC it does list the amount of courses that you’ve taken in each subject and your average grade in that subject. What’s more important than your high school grades is your score in the matriculation examination, a national-level standardized test that’s held in the various subjects. There too you have considerable freedom in choosing which subjects to take the exam for, and on what level. When you’ll apply to higher education later on, the different universities and polytechnics will each have their own criteria for how much weight they give your scores in any given field – if you only got tested in short math and barely passed it at that, that might not even be a problem if you apply to study in some program which values something else much more than good performance in math.

          Also, not everyone goes to high school. After elementary school, you can basically choose to pursue either vocational school or high school, where vocational school is intended to give you an occupation and high school is intended to give you a well-rounded basic education and prepare you for university. In principle, these do not select on ability, but the schools do get to choose who to admit based on the students’ elementary school grades, and everyone knows that high schools tend to require a higher average grade to admit you than the vocational schools do. So there’s selection at that stage as well – though again, the student is free to apply to any school that they wish, and there are some vocational schools that require a high average grade as well as some high schools that will let you in with a low one, leaving enough wiggle room that people can ignore the fact that there’s ability-based selection going on.

          None of this really applies on the elementary school level, which is basically the same for everyone (unless you have disabilities), though.

        2. Anthony

          You can have classrooms called A and B, you are just not allowed to openly speak about the fact that A is an elite classroom, and B is where you put the rest.

          Even if the teachers and administrators don’t talk about it, the kids will notice. And when they grow up, they’ll know which class their kids got put into if they pay any attention at all.

          You could probably improve your odds of getting away with it if you mixed things at the borders, especially if you made sure that kids near the border got to spend some years in each class.

      2. Kevin

        people will get angry if children are stratified based on ability

        I think there’s a lot more variation in ability grouping across the US than people recognize. Let me describe my own experiences with it, in the public school system of a relatively small city.

        Elementary school: we had “reading groups”, “spelling groups”, and “math groups”. We also had a decent gifted and talented program. Nobody really raised much of a fuss over these things, as far as I can recall. There was a lot of teacher discretion in how they were implemented. For example, in fourth grade, I was put in my own spelling group, and eventually no longer given spelling tests at all once my teacher realized that I could spell any word in the dictionary.

        Middle school: the principal was highly opposed to ability grouping for unspecified reasons. (My experiences with this man have convinced me that most education reform discussions underestimate the impact that poor administration can have on a school.) Parent opposition to the no-ability-grouping policy managed to eke out an honors math program, as well as a (disorganized, vague, mostly useless) gifted and talented program.

        The presence of gifted and talented programs does tend to invite the phenomenon of parents pushing for their kids to be included in the program even when it’s not appropriate or useful for the kids. My mother always told such people that they should think of the gifted and talented program as another form of special education, which is technically true but ignores status-based motivations.

        High school: there were three “tracks”, basically remedial, standard, and honors. There was also an accelerated math program beyond the standard honors courses, and there were some AP courses available. Incoming freshmen were placed in a “track” for a given subject based on their middle school record and teachers’ recommendations. I’m using “tracks” with scare quotes because, after the initial placement in freshman year, course levels were mostly up to the students. If you did well enough (B average or higher) in a course, and got any teacher or guidance counselor to sign a form, you could move up to the harder “track”. If you wanted to move to an easier “track”, I think you could just choose to register for the appropriate course.

        We did have to deal with the local Green Party trying to do away with ability grouping through various underhanded tactics, but we were able to fend them off. The local newspaper consistently misrepresented the extent to which students were forced to stay in a certain “track”, and they never printed any of my letters to the editor correcting them.

        There’s an interesting discussion further up in the comments about the empirical efficacy of ability grouping (although data seem somewhat scarce), but I’m sure that people’s anger has little to do with empiricism. I suspect it’s much more based on parental status: parents are the ones perpetuating ideas of merit-based personal worth, and then getting upset about them when they aren’t winning on that metric. Kids will insult each other about being smart or stupid, but that happens with or without ability grouping. Kids basically know how well their peers are doing in class.

        Most kids want a certain amount of work and a certain pace for a given subject. Forcing them to a different amount of work or pace has never seemed like a sensible strategy to me. [insert Harrison Bergeron reference]

        Anyway, long comment is long. My point is that both ability grouping systems and people’s reactions to them can have a large amount of variation.

  30. Charlie

    So finally out of desperation I was like “Who wants to do some kind of idiotic activity in which we all pick English words and color them in and then do a stupid dance about them??!”

    From one perspective, sure it sounds bad if you have to “trick” them into learning.

    From another perspective, though, your average students don’t actually care about the material, and the idiotic activities are simply contexts where they’ll learn automatically. It is not a moral failing by the students to learn best in this context.

    Though of course, there are more-advance teaching strategies that can circumvent some of these limitations, like cooperative learning. From an article: “The group mostly consists of a fast learner, a students with learning difficulties, and two average students.” Why is the group like this? Why yoke a fast learner to a slow learner?

    Because each student is there to do a different thing. The fast learner is there to be exposed to the complicated parts of the project and explain the hard parts to the other students. The average students are there to do the easier parts of project and get rote exposure to the material. The slow learner is there to get the most passive exposure to the material.

    And so you codify these roles – you give each student a role that emphasizes what they should be doing, and you make those roles meaningful by restricting the other students. One student is the only student allowed to write things down – the Recorder. What are they practicing? Memorization and general structure of the material. Typical average kid task. Etc.

    And this successful teaching technique looks almost nothing at all like “say to them what I think I would have found interesting.” Which we could have antipredicted, based on how weird an environment school is relative to spontaneous conversation.

    1. David Gerard

      Frankly he should just go for it IMO. Particularly with his interest in LessWrong – there’s little that’s quite as fascinating as watching a small intelligence grow, and seeing all the hundreds of abilities we so blithely blanket-label as “intelligence”.

      I must admit, my worries as a new father were greatly calmed by my loved one already having two daughters, so being quite experienced 🙂

      1. Anthony

        there’s little that’s quite as fascinating as watching a small intelligence grow, and seeing all the hundreds of abilities we so blithely blanket-label as “intelligence”.

        Oh God yes! My two daughters (4 and 6½) are figuring out all sorts of things, and at fairly different paces from each other. It is fascinating.

  31. Viliam Búr

    Being a former teacher, thank you so much for writing the third part! I would love to see all those internet experts on teaching go to a real school with real students, and see their perfect plans shattered again and again, predictably. Or at least to realize that teaching in one classroom can be a very different experience from teaching in another classroom, and how completely differently even the same class behaves on days when the one specific child is sick at home.

    The teachers are judged by results that for the most part don’t depend on them. Even the individual student’s results for the most part depend on how their classmates behave. This is completely ignored in most discussions about education. I wonder whether this will be improved by an online education, when the students will less depend on their classmates’ cooperation during the lessons.

    You also mentioned child psychopaths: in a sufficiently large school, a few of them will appear predictably. Yet it is a big taboo for a teacher to even hint at this possibility. So obviously, when such students predictably appear, a typical school is completely unprepared. Usually the teacher is blamed for not magically solving the problem.

    1. Anonymous

      Depending on the location, you might be able to filter out all but very high functioning sociopaths out of even a very large school. Where I went to high school, for instance, there were no students who were *too* horrible because anyone with social difficulties of that level would have been sent to another school nearby for kids who weren’t ready to be “mainstreamed.”

      My school had the motto “A tradition of excellence,” which I used to laugh at, both because it was so generic that it clearly required no attention to any qualities that might have been particular to that school, and because I thought it was so off-base, since we had all the mindless rote repetition, vapid activities, and failure to instill any sense of purpose or pleasure in learning, etc.

      It wasn’t until after I graduated that I realized just how much worse it easily could have been, and usually is at other schools. The students who won national awards and went on to be hedge fund managers and such might have been learning mainly on their own initiative, but at least the environment didn’t inhibit them to the point of making those achievements nigh-impossible.

  32. Throwaway

    I think nowadays the possibility of severe autism may worry people more than the Down syndrome and the much rarer ones.

    I have a 5 y.o. daughter on the autistic spectrum. Three years ago when we got the diagnosis, she had no language, having regressed away from a few initial words she’d had, and it was not at all clear she would ever speak. Let me tell you, it was not a “I should just gaze deeply into her eyes” sort of situation for me. I did love her unconditionally, but I was desperate and heartbroken and I would sometimes lie at night and sob uncontrollably at just the thought of her not growing mentally.

    Things are much, much better now (calculus at age 5 is not in the cards, but her language and raw intelligence have caught up with her age, social interaction is the remaining major problem), but since she’s still in the special needs kindergarten system, I sometimes see and play with some of her classmates who didn’t start speaking. I honestly don’t know how their parents cope; every time I play with one of them, later at home I have a fit of rage and despair and can’t calm down for a few days afterwards. I don’t know how you cope in the hospital either, come to think of it.

  33. g

    I’m just going to agree with the other people saying: Yes, if you have children there’s a small but non-negligible risk that it will make your life really awful, but likewise if you make a new friend or take a new job or just cross the road or have a meal there’s a non-negligible risk that it will make your life really awful.

    (New friend might turn out to be a serial killer and botch your killing, leaving you with lifelong disfigurement and mental impairment. New job might turn out to be at a place that’s a front for something very evil, with suitably evil ways to stop you leaving once you’re there. Crossing the road might result in being hit by a car, with all the lovely consequences that can entail. Food might happen to be contaminated with disastrously poisonous compounds of mercury or aluminium or something.)

    You really don’t have the option of avoiding all risks of terrible things. (Just doing nothing carries its own risks.) So you weigh up the costs and benefits and do things for which the net benefit seems like it outweighs the risks and other costs. If you really want children and your partner really wants children, the benefit may well outweigh the risks and costs for you. That, not “is the risk non-zero?”, is the question to answer.

    1. Scott Alexander Post author

      I am aware of the existence of risk-benefit calculations. Sometimes they’re just hard.

  34. David Gerard

    Would donating sperm satisfy the urge to breed? You’re smart, you’re a doctor, your sperm are golden from that angle. If your sperm are biologically up to the task, it’s pretty much certain they’ll be used, lots and lots. And the children will be raised by people who very much wanted them and will not fuck them up.

    1. Creutzer

      And the children will be raised by people who very much wanted them and will not fuck them up.

      What are your grounds for the latter assumption, and for the implicated claim that there is a significant correlation between wanting children and not fucking them up?

      1. Scott

        Even if sperm-donor parents are no different in how much they want children (and how much time, effort, and money the are willing to expend), it’s not like you accidentally stumbled into a sperm bank and got impregnated before you realised what was going on. So accidental pregnancy for unprepared parents is excluded.

        It’s no stretch to think that unprepared parents, possibly without the financial wherewithal to support a new child, and harbouring some latent misdirected resentment for the child interrupting their lifestyle, are going to be worse than average parents.

        I think it’s pretty obvious that if you exclude a bunch of “on average, worse” parents, that decreases the likelihood of the child being fucked up.

        (This is the same argument that I make about homosexual couples having children – it’s actually on average better than a heterosexual couple having a child, since you can’t have an accidental pregnancy in a same-sex relationship.)

        1. Creutzer

          I totally believe that. But that does not warrant the assertion that a child raised by people who make use of sperm donation will not be fucked up. It’s very much possible for well-meaning and financially secure parents who fully intend to have a child mess it (or whatever pronouns you’re supposed to use to refer to a child) up. And especially if the sperm is used “lots and lots”, chances are that not all of the children will come out alright.

          1. David Gerard

            Yes, I meant “less likely to be, given they’re definitely wanted.” as Scott notes. (I was adopted myself, so I have personal experience of having been thrown to the winds of fortune and how that worked out.) Sorry for the overstatement.

        2. Mary

          On the other hand, people have higher expectations for things they wanted and planned for. And when the child fails to live up to expectations, the danger of abuse arises.

      1. David Gerard

        Not a one. I’m slightly wishing I’d done it through a commercial sperm bank than an NHS hospital, because the latter is stultifyingly bureaucratic, and the former would probably spread my genes further. But hey, I’m past 46 and no longer eligible to donate.

    2. Mary

      Actually abused children are more likely to be not only wanted but planned than a control group. Their mothers, for instance, went into maternity clothes earlier.

      After all, the higher your desires for the child, the more likely you are to have a child who doesn’t fulfill them, and that’s a point of danger.

      1. David Gerard

        I do also have a daughter I’m raising (and it’s absolutely amazing and I wouldn’t miss out on it for anything), so I appreciate that they’re very different things. However, I would suggest you consider it as an additional thing.

  35. Chris

    life without risk is not living.

    have you ever been on board a moving vehicle? similar long-term-quality-of-life altering risk (although the risks are probably greater than those you outline above).

    1. Elizabeth

      Life without risk is wildly, overpoweringly preferable to life with risk. What’s being discussed here is the chance that a parent might not be able to take adequate care of their child. That’s not small potatoes.
      Yes, the rest of life also contains small chances of terrible outcomes. That’s not a reason to be less afraid to reproduce, I think it’s a reason to be more afraid (/cautious/pessimistic/thoughtful) about everything.

        1. Randy M

          Well, they do if you are a driver in a car with passengers or on a street with other people, etc. There are quadriplegics* due to accidents no fault of their own, and there are people who accidently start fires or spread disease through no fault of their own. Not many, but nor are there many of the worst cases Scott fears from ‘genetic Rousian roulette.’

          *That word has a lot fewer ‘a’s than I thought before spellchecking.

        2. houseboatonstyx

          @ Randy
          As a driver, you risk being sued for everything you have, with judgments against you for whatever you may get in future. But even in that worst case, you don’t become legally responsible to care for and protect the victim all zis life, with any other needs he may have. Which likely includes having him in your home, doing 24/7 personal care, bailing him out of jail, etc etc. (And the risk that if the victim decides to harm you, he has full access.)

          The only way you risk this liability for an accident victim — is if the victim is your own (till then perhaps normal) child.

          This kind of legal (and perhaps moral) responsibility is a whole nother level of horror — and is avoidable.

        3. Randy M

          I was responding to von Kalifornen, who mentioned emotional risk. But reading more carefully I see it says the emotional costs of fears of responsibility. So, yes, it is unique in that way.

  36. suntzuanime

    You might end up with a terrible child for no reason, but you probably won’t, and what else were you going to do with your life? I mean, you’re part of the LW cluster of folks who are much more likely than average to have Grand Quests that a problem child would derail, but for most people raising children is by far the most Meaningful thing they do with their lives.

    Even if you don’t do anything else useful, if you raise two children with as much potential as you had, you break even and you get a do-over in the next generation. At three children you’re actually profiting. Having children is a powerful force for imposing yourself on the world.

    1. Vanzetti

      >>>but for most people raising children is by far the most Meaningful thing they do with their lives.

      Meaning has nothing to do with. Every fucking living thing (pun intended) on the planet makes children. It’s as meaningful as gravity. You don’t get to pick a meaning for obeying gravity.

        1. DSimon

          That’s an amazingly cool statement. Thank you very much for it. 🙂

          Sometimes when I’m explaining why evolution is not a system of morality even though it a significant part of how our current system of morality came about, I use a gravity metaphor. I say:

          Evolution explains why we value cooperation, sure, but so does physics. Evolution is implemented on physics as a platform, and if, say, the gravitational constant were different, then we could well have turned out differently.

          But we do not have any moral obligation to push people down stairwells, even though that’s what gravity “wants”. Instead, we get to come up with our own values for our lives.

          The reason I bring this all up is because now, thanks to you, I have a fun positive way to end this spiel, instead of leaving my audience with a mental image of someone being pushed down stairs. 🙂

        2. Vanzetti

          >>>But we do not have any moral obligation to push people down stairwells, even though that’s what gravity “wants”. Instead, we get to come up with our own values for our lives.

          Physics is also what makes you push or not push someone down the stairs. Where’s the morality? 🙂

  37. roystgnr

    “no idea”? You haven’t even tried a back-of-the-napkin expected utility calculation? I know that trying to use math with hokey numbers isn’t a great way to make decisions, but at least it lets you try to get a less biased viewpoint. If your job buries you in a population of people selected for their difficult mental problems, you can’t expect your intuition to keep telling you anything useful about how much you should worry about mental problems among people (like your own hypothetical kids) who haven’t been through such massive selection bias.

    IIRC the research says that having kids makes your own life feel less “happy” and more “meaningful”, but you seem to be altruistic enough that that tradeoff probably wouldn’t be a deciding factor for you. From the altruists’ standpoint I guess the pertinent question is “does the world need more of the kind of people that my kids are statistically likely to become”, and in your case the answer is pretty clearly yes.

    1. jooyous

      It’s one of those things where the odds of something horrible happening are small, but the horribleness is horrible enough that it’s scary to take them anyway.

      I also feel like it’s hard to calibrate between the selection effect of scary mental problems and the narcissistic trap of “well MY kids won’t be like this”. I think most people, at least in my parents’ generation didn’t stop to wonder about this at all. I don’t think it even crossed their mind that their kid might be on trial for murder with KILLER written to his shirt. How many families with unfixable kids have you met or heard of in your lifetime? How many is average? How are those families doing?

      1. Randy M

        “It’s one of those things where the odds of something horrible happening are small, but the horribleness is horrible enough that it’s scary to take them anyway. ”

        Of course, one can say the same thing about walking into a hospital. Life is risk. Making life all the moreso.

        1. houseboatonstyx

          All life is risk — which we cannot completely avoid.

          Making life — is a risk which we can completely avoid.

        2. Randy M

          I supose my closing comment there was a bit of irrationality–since a certain amount of risk is unavoidable, then we should be comfortable with a level of risk elsewhere of comprable level, even if avoidable.

    2. Cyan

      If your job buries you in a population of people selected for their difficult mental problems, you can’t expect your intuition to keep telling you anything useful about how much you should worry about mental problems among people (like your own hypothetical kids) who haven’t been through such massive selection bias.

      Quoted for truth.

      At a LW meetup last year, a buddy of mine told us he was going in for surgery to remove part of an over-active thyroid gland. He was experiencing intrusive thoughts about the possible bad outcomes: the risk of losing his voice was about 1 in 100; the risk of death was about 1 in 10000.

      In contrast, his wife is a nurse who preps patients for surgery all the time, and she was totally unconcerned. This is an inversion of your situation — her job trained her intuition to expect the right prevalence of surgical complications for this procedure.

      I collected 13 pennies from people at the table (this was back when we had pennies). I told him that I would toss them to simulate his surgery — the probability of his dying was less than the probability of all of the coins landing heads. Then I ran the simulation seven or eight times, and he “survived” in each one — after which he felt much better!

      1. MugaSofer

        That’s a pretty good idea, actually.

        Although ironically, I would be too scared they would all come up heads by chance (and terrify the poor guy) to do it :p

    3. Scott Alexander Post author

      Down Syndrome is 1/700, autism is 1/100, and cerebral palsy is 1/500, for a total chance of 1.5% chance of one of these serious neurological disorders; I’m going to round up to 2% to add in all the little ones.

      Schizophrenia is 1/100, and antisocial personality is 2/100, for a total chance of 3% chance there. I worry I might have an increased risk here because I’m dating Ozy who has borderline personality disorder and relatives of people with that disorder have a 10% risk of antisocial personality disorder.

      So the average person’s chance of having one kid with one of those five disorders is 5%. My chance might be as high as 13%. Increase the number of kids and I increase the chance. I don’t think it’s as small as everyone seems to think.

      1. AndekN

        These numbers don’t take into account prenatal screening. If you are are ready to abort in case of positive result, then the actual probability of having a child with a syndrome should drop to less than 1%.

        1. coffeespoons

          Agreed. However, no screening exists for Anti-Social personality disorder, and 10% does sound alarmingly high.

        2. Scott Alexander Post author

          As far as I know, of the conditions listed only Down Syndrome has anything close to a screening test.

      2. Crimson Wool

        I worry I might have an increased risk here because I’m dating Ozy who has borderline personality disorder and relatives of people with that disorder have a 10% risk of antisocial personality disorder.

        So find some other woman to have kids with, or get an egg donor.

      3. ozymandias

        BPD and APD are both correlated with various kinds of absolutely horribly bad parenting, so I suspect the true genetic risk is much lower than 10% as long as we refrain from child rape, attacking our children with beer bottles, etc.

        1. Scott Alexander Post author

          Reasonable unless you believe the whole behavioral genetics thing where parenting really doesn’t do much.

          Although it’s possible their studies just failed to capture parents sufficiently bad to make a difference, I’m surprised that bad parenting isn’t dose dependent and I feel like assuming it would be kind of motivated.

        2. ozymandias

          Even so the 50% that’s unshared environment probably matters (and in one of Judith Rich Harris’s books she suggests that some of unshared environment *is* parenting– people don’t treat their children the same). In particular, the rate of history of childhood sexual abuse among borderline patients is extraordinarily high and it seems plausible that’s a causal factor for BPD. It also seems plausible that families with children who were raped as children are at higher risk of children suffering other traumas.

          In addition it seems prima facie implausible, given that we know abusive relationships in adulthood fuck people up, that abusive parental relationships *wouldn’t*.

        3. Randy M

          “Reasonable unless you believe the whole behavioral genetics thing where parenting really doesn’t do much. ”

          Seems at the least it would be like Iodine, right? More won’t help, but if you are totally lacking you’ll suffer. But I’m not conversant with the studies you reference.

      4. Michael Edward Vassar

        FWIW, my impression is that all of those things except for cerebral palsy and (probably) Down Syndrome are preferable to neurotypicality combined with contemporary culture and IQ between 115 and 130.
        Look at
        and the nearby links wrt antisocial personality disorder. It can be terrible, but can also be a reasonable life. Schizophrenia is likewise frequently downright awful, and autism can be, but one of the two, well managed, seems nearly necessary for a really intellectually productive life.

        1. Anonymous

          Well, if you define “preferable” solely by intellectual productivity. Most would prefer an unintellectual child to a sociopathic one incapable of feeling empathy.

      5. Cyan

        I’m not saying that I think the prevalence of bad outcomes is as low as 1 in 10000. I’m saying that your subconscious future-experience-anticipator unit has probably been affected by selection bias, because in your position you see the worst of it. Further, there’s an easy method to at least try to retrain your intuition: give it a whack-load of data with the right frequencies. (Perhaps spend a bit of time after each simulation visualizing the simulated outcome.)

        Of course you may still find the bad outcomes outweigh the good ones in expected utility. My point is more like: if that’s the judgment at which you arrive, I would expect it to feel a whole lot less fraught and a whole lot more well-founded.

        1. houseboatonstyx

          I’m saying that your subconscious future-experience-anticipator unit has probably been affected by selection bias, because in your position you see the worst of it.

          There may be selection bias on the other side, too. Even by the lower figure, 5%, five* out of 100 parents would be caring for permanently-helpless offspring needing special care, special facilities, etc. But are those the parents that most of us meet, or the children that most of us see? Or are those parents more likely than average to be stay-at-homes, and those children more likely than average to spend time in some special care facility instead of easily visible, etc?

          * Actually, if there is a 5% chance per birth, then 10% of parents would be affected. That’s just for single-child families.

          So the intuition that brushes off the danger as too rare, may be formed by experience in a group already filtered.

      6. DavidS

        I don’t know all those disorders, but autism and ASPD have a very wide spectrum. Since a large portion of your estimate comes from dating Ozy, I think you should ask yourself “Would I raise a child like Ozy was as a child? Of that 10%, how many are more severe than their parents?”

  38. Alyssa Vance

    “But I would still be unexpectedly finding out that what I thought was just a perfectly normal super-gargantuan committment to sink most of my resources into a mostly-helpless creature for twenty-five years would in fact become, without my planning or consent, a super-ultra-titano-gargantuan commitment to sink even more of my resources into an even-more-helpless creature indefinitely.”

    You know what, I object. I object to the notion that, in order to ever have children, one must dedicate *most* of *any size* pile of resources to them for twenty-five years. That seems silly, ahistorical, and (for a sufficiently large pile) probably harmful to the child. (See eg. for a book by someone smart who mostly agrees with this view.)

    If you have kids, I’d seriously look into leaving the US. There is demand for doctors pretty much everywhere, and the US government does a *lot* of trying to use ‘for the children!’ to impose restrictions on parents (of which trying to increase the cost of everything kid-related is only one component). It is, for example, considered perfectly normal to use children to legally prohibit someone from having sex ( This only came to people’s attention when it happened to be gay sex and was therefore noticed by the gay rights movement.

    1. Said Achmiz

      While the background propositions that go into this view of parenting seem essentially correct to me, I remain highly skeptical of it as an actual approach to parenting until I see some testimony from actual parents who went into the childrearing experience with this view, raised some children while following its recommendations, and have remained adherents. (That may be an implausible demand, given that it’s a relatively recent view, so I might provisionally accept testimony from parents who are went in with that view and are partway through the process of childrearing while remaining adherents.)

      In other words: this seems like one of those plans that would not survive contact with the enemy.*

      *More so than most.

      1. Randy M

        I’m not sure what that approach actually is?

        I find children can cost less $ than expected, especially if some convenience items are eschewed. But they take time and attention, most of one persons most of every day.
        The marginal increase isn’t terribly much after the first couple years, barring the types of situations Scott describes.

    2. Scott Alexander Post author

      Although I agree with Caplan’s thesis, a few caveats:

      – I think there’s very good evidence that parental attention is very important for babies. For example, babies breastfed on demand end up with higher IQ than babies breastfed when it’s convenient for parents, and last time I checked it didn’t look like this was obviously just confounders.

      – A large part of the investment in child-raising is structural – getting them to and from school (or home-schooling them), getting them food and clothing, et cetera.

      – Besides the “invest time in your kids because it will help their neurodevelopment and make them smart” argument, there’s still a lot to be said for doing it in order to be a halfway decent person. If your kid is having a terrible time at school and comes home crying, there’s a lot of wiggle room in what exactly you should do, but I think most people would agree that telling zir you don’t have time to listen and you don’t want to hear about it isn’t really a going option. That commits you to a lot of emotional investment.

      – Caplan’s argument doesn’t decrease financial costs much, many of which are things like bigger house, more food, clothing, and (especially) college tuition if you plan to help with that (which I would; my parents helped with mine, it saved my life, and I want to pay it forward)

      1. Michael Edward Vassar

        If your kid is at school, given current legal structures making other options available, you’re already not a halfway decent person.
        Parents all *sometimes* tell their kids that they don’t want to hear about it. There’s no alternative.

        1. Randy M

          I lean towards your views, I think, but are you really comfortable with a conclusion that the vast majority of parents currently are not half-way decent people? Do you include private school in that?

        2. Said Achmiz

          Going to have to disagree with you there, Michael. I am having a hard time imagining anything better my mother could have done for me than have me attend the high school I did. Maybe this doesn’t satisfy your “other options available” stipulation? What other options are you talking about, anyway?

  39. Alyssa Vance

    “Schools are evil fascist institutions that deliberately avoid teaching students for sinister reasons.”

    I believe your observations, and *also* agree with this. The problems start way, way, way further up the chain. If you forcibly round up a bunch of eight-year-olds, plonk them all down in a room, force them to sit still for an hour, and then force them to ‘learn’ a bunch of mostly useless things chosen mostly to please whichever politicians are currently in power, then of course you’re going to have problems. Compare to, eg. Sudbury schools (

    1. Keller

      Alyssa, I don’t think that your example supports your claim. What we have is a school system that simply doesn’t have “actually teaching children” at the top of its list of priorities. Your example supports that argument, because the facts we want to teach children are dictated by elected officials and that is many kinds of fucked up. However, you haven’t shown that there is any attempt to “not teach children”, just that we value other things way more than we do “actually teaching children”.

      1. Randy M

        “Your example supports that argument, because the facts we want to teach children are dictated by elected officials and that is many kinds of fucked up.”
        Curious as to your alternative suggested method. (I have better ideas myself, but interested from which direction you are attacking the curriculum selection methods.)

      2. Michael Edward Vassar

        Alyssa points out a school that explicitly doesn’t teach children but which gets better results than our schools do WRT their abilities at the things our schools supposedly teach. is another example and there are many more.
        When an intervention doesn’t systematically tend to produce the effect of ‘children being taught’ it’s fair to say that the intervention is not ‘teaching children’

    2. Scott Alexander Post author

      Let me specify my claim, then. Given:

      1. An environment with about thirty students per teacher
      2. Students unselected for intelligence or good behavior
      3. Teachers who have to be mass-produced so there are enough of them for the whole country
      4. No more budgetary resources than the average public school
      5. Parents demanding their kids be supervised for eight hours/day so they can go to work
      6. Someone will get angry if you stratify students by ability and say it’s a violation of the fundamental dignity of the human person
      7. The demand that a large percent of students (no fewer than in current system) learn reading, writing, arithmetic, and very basic science like that things are made out of atoms, and very basic history like that the Civil War was a thing that happened…

      …it will be difficult to systematically outperform the current system. Do you agree or disagree?

      1. Crimson Wool

        6. Someone will get angry if you stratify students by ability and say it’s a violation of the fundamental dignity of the human person

        I recommend we get rid of these people somehow, possibly by putting them into a rocket and firing it into the sun.

        1. AndekN

          Nah, the real reason against stratification has nothing to do with dignity of a person. It is much simpler: stratification doesn’t work.

          Sure, the highest group performs better after ability grouping, but the middle and the lowest group perform worse (Cahan et al 1996). As total utilitarians, this is not a result we should strive towards.

          And it gets worse. Although the highest group gets better results, at the same time their academic self-image gets worse (Linchevski & Kutscher 1998) and they end up applying for higher education less often than their peers in non-stratified classes (Mulkey et al 2005). This is probably due to students comparing themselves against other students in the same class. If you are the worst of your class, you will feel like a loser, regardless of how good the other kids at your class are.

          You can have differentiated teaching in the same class, and you will get better results that way (Kulik & Kulik 1992). But creating different classes (or entire schools!) based on ability leads to poor consequences.

          Cahan, S., Linchevski, L., Ygra, N., Danziger, I. (1996). The cumulative effect of ability grouping on mathematical achievement: a longitudial perspective. Studies in Educational Evaluation, 22:1
          Mulkey, L. M., Catsambis, S., Steelman, L. C. & Crain, R. L. (2005). The long-term effects of ability grouping in mathematics, Social Psychology of Education 8
          Kulik, C. L. & Kulik, J. A. (1992). Meta-analytic findings on grouping programmes. The Gifted Child Quarterly, 36:2
          Linchevski, L. & Kutscher, B. (1998). Tell Me with Whom You’re Learning, and I’ll Tell You How Much You’ve Learned: Mixed-Ability versus Same-Ability Grouping in Mathematics, Journal for Research in Mathematics Education 29:5

        2. Crimson Wool

          That, too, would be a violation of the fundamental dignity of the human person.

          Have you considered that it might be a greater violation of the fundamental dignity of the human person if we don’t do it?
          @AndekN: I’ve read on the subject previously (during my brief flirtation with becoming a mathematics teacher), and the evidence I read suggests a radically different position from yours (e.g. Kulik 1993 seems to pretty clearly argue in favor of tracking). I will be reviewing mine and yours and deciding what to believe.
          However, from some initial checking, it seems your reading of Kulik 1992 is based on a comparison of just XYZ and Within-Class grouping, ignoring Cross-Grade, Enriched and Accelerated grouping styles (all of which substantially outperform un-tracked classrooms). XYZ grouping is inane, so I don’t think that’s a fair comparison. Accelerated grouping wildly outperforms all other grouping styles.

          This is probably due to students comparing themselves against other students in the same class. If you are the worst of your class, you will feel like a loser, regardless of how good the other kids at your class are.

          Assuming it’s real, what would the logical extenuation of this effect be for home schooled students? I mean, obviously, home schooling supremacy, but still.

        3. AndekN

          it seems your reading of Kulik 1992 is based on a comparison of just XYZ and Within-Class grouping, ignoring Cross-Grade, Enriched and Accelerated grouping styles

          I really simplified too much in my comment, and you are right: the case against ability groups is not as clear-cut as I made it to be. Cross-grade grouping showed mainly positive results for all levels, as you say. However, in practice XYZ grouping seems to be more prevalent, and I think that when people are imagining ability classes, they are imagining XYZ groups. Perhaps I’m mistaken, but this is my impression after numerous discussions with teachers and non-teachers alike.

          Studies about Accelerated and Enriched (A & E) classes are interesting, but they often ignore the effect of A&E classes to the schools as a whole. If the gifted children study in A&E classes, then the rest of the classes are composed of middle- and low-performing children. I haven’t seen research as to what happens to those classes, but the studies I cited would at least hint to the conclusion that their performance would drop. If you have sources to the contrary, I would be very interested in hearing about them.

          Assuming it’s real, what would the logical extenuation of this effect be for home schooled students? 

          I have no idea. They are the best AND worst of their “class”? I would perhaps expect something like Dunning-Kruger effect: they would overestimate their ability if they are weak and underestimate it if they are strong. Maybe?

      2. Douglas Knight

        As far as I can tell, simply eliminating instruction satisfies all your points.

        It meets 6 and 7, but people will get angry about them out of status quo bias and because the current system is designed to obfuscate them.

        The main deviation from the current system are (1) reducing suffering and (2) failing to extinguish the natural love of learning. From my point of view, those are the main upsides. From the point of view of society, it fails the main requirements. Also, the students are extremely unconfident about math.

        1. Scott Alexander Post author

          I’m unclear how school minus instruction would be different from a child’s free time, which many of them spending watching TV and few of them use teaching themselves algebra. Why would you expect this to change when they’re in a school building?

        2. Douglas Knight

          I expect it to work in the future because it worked in the past. As I said in my first comment, this is not hypothetical.

          No, the school building doesn’t do much. Children are just as curious at home. Centralization at school is good if the children want to hire tutors and spread the cost.

        3. Alexander Stanislaw

          But has it worked w.r.t. the demographics Scott is concerned about? If you’re talking about the Sudbury Valley School, then getting white middle-upper class students to succeed is nothing to marvel at. Maybe they did better than they would have otherwise, but they were going to succeed in any halfway decent educational system. Its getting an education system to work for the left part of the bell curve that’s the tricky bit. Has it been tried with more difficult to teach demographics?

        4. Michael Edward Vassar

          @Alexander Stansialw,
          There used to be a Summerhill style Harriet Tubman Free School in rural New York aimed at rural African American youth with, like all these efforts I have ever heard of, surprisingly good results.

      3. Michael Edward Vassar

        I strongly disagree with this claim. Are those the only constraints? Do you know what the budgetary resources of a public school *are*?

        Actually, I’m open to the hypothesis that I personally would find it very difficult to do anything with budgetary resources comparable to those of the US school system, because such resources, many hundreds of billions of dollars, are very difficult to dispose of in a manner that is beneficial at all on the margin. I’m not entirely sure that I know people who can do things at that scale successfully, WRT things in general, but I’m extremely confident that I could produce MUCH better results than .01% of the public school system with a random .01% of the kids and of the funding.

        1. Scott Alexander Post author

          Well, yes, if you only had to scale to a size of 0.1%, you could just nab some really good teachers. Can we change your scenario so that you need to gather a pool of several million teachers, and we randomly select 0.1% of them? And same with administrators?

        2. Michael Edward Vassar

          I’m not assuming that I have access to more than .1% of anything. Just give me a random bit of the US. I’m also not assuming that I want to use ‘school teachers’ at all. I expect I’d use a mix of computer instruction, video games, old people, teenagers skilled tradesmen, small business people and grad students. Maybe even some military officers.

        3. Michael Edward Vassar

          @ gabriel,
          what I have in mind involves less focus on computers, more mentoring and independent exploration, but I expect your proposal would work too.

        4. Gabriel

          I do wish you would write up your proposal (and generally do a lot of blogging).

          Also, St. Rev is responsible for that post. I merely requested it.

      4. St. Rev

        Just as a reality check here, contrast institutional ‘music education’ to self-education. Kids absorb astonishing amounts of information about their favorite bands; many of them start their own. Meanwhile, music education in the school system gives you marching bands, grotesque militaristic pantomimes of interest to no one in the wider culture.

        1. Scott Alexander Post author

          Agreed, but music self-education isn’t consistent. That is, have no formal enforced music program in schools, and some percent of kids (10%?) will form bands and self-educate to some level of proficiency with an instrument. And 90% won’t.

          As long as the demand is that the population as a whole be trained in something (my point 7 above) this isn’t politically feasible.

          This segues into really good questions about what the point of the school system is, and I don’t really know and I don’t think anyone else does and it could be that it has no point. But as long as the point of the school system is what it is (every child must be trained in a certain set of skills including useless ones which they hate) things like music self-education aren’t going to cut it.

          I’m arguing that holding the politics and administrative demands constant, there’s not much low-hanging fruit an individual teacher can do take to be better than current teachers.

    1. Kaminiwa

      The link doesn’t speak to problem children at all. Does this school, being private, actually accept problem children in the first place? Does it have a track record of turning around said previously-problematic kids, or does it just expel them?

      If there’s a school that can routinely fix “wants to vivisect cats”, I’d be a bit surprised 🙂

  40. Ozy

    Aren’t we getting to the point where early experiments in genetic engineering should be available to would-be parents dedicated to getting it, for sufficiently extreme definitions of “dedicated?”

    Arguably, we already have that, if you include defect-selective abortion. That at least covers the first category of your concerns, right?

    1. coffeespoons

      Not entirely. I know someone who has a kid with a genetic syndrome. It didn’t show up in and of the pre-birth screening. They had no idea that their child wasn’t totally healthy until he was born.

    2. MugaSofer

      I think Scot isn’t worried about this – and you would know better than I do – but the potential ethical issues with that sem worth treating it as a separate category. (After all, human GM advocates probably don’t want “kill the impure” associated with the concept for obvious practical reasons, if nothing else.)

    3. Mary

      No test is infalliable. Indeed, there’s no test for which the false positive rate is lower than the incidence rate of the disease it’s testing for, so you are much more likely to have a healthy baby than a sick one EVEN IF the test says the baby is sick.

      Not to mention false negatives.

    4. Scott Alexander Post author

      Not yet. Lesch-Nyhan is detectable but 1/3 of cases are de novo mutation (ie the parents don’t have it but the kids do). And we’re still in the infancy of understanding psychiatric genetics.

  41. Leah

    This is also my biggest fear (and the cause of my break up with my first boyfriend, when discussed only hypothetically). I’ve got bupkis in the way of solutions.

      1. Jeff Kaufman

        a) You can’t generally tell if a kid has no problems before adopting them. Primarily because some problems won’t be visible until they’re older.

        b) The pool you’re drawing from has a higher incidence of problems.

        I would expect this strategy to be less likely to end up with a low-problems kid.

  42. Karl Gallagher

    Evolution is not a spectator sport. If you want the next generation to be more like you, have kids. If you want the next generation to be less like you, don’t have kids.

    1. Vanzetti

      >>>Evolution is not a spectator sport. If you want the next generation to be more like you, have kids. If you want the next generation to be less like you, don’t have kids.

      Or encourage people like you to have more kids. Or neuter people who are less like you. Or use genetic engineering.

      Or realize that wanting *anything* to happen after your death is utterly irrational.

      1. Karl Gallagher

        Or encourage people like you to have more kids.

        Example seems to be the most effective method for that from observed data.

        Or neuter people who are less like you.

        That’s a felony last I checked.

        Or use genetic engineering.

        Tampering with other people’s kids would also be a felony.

        1. Randy M

          Well, from a purely utilitarian standpoint, things that are felonies tend to for that reason be hard to do successfully.

      2. Steve

        To expand on roystgnr’s objection to your use of “irrational”: Would you accept $1, an hour before your death, in exchange for all your friends and loved ones being tortured indefinitely, an hour after your death?

        1. Anonymous

          Vanzetti, you seem to be using a definition of “rational” that implies egoism, which is philosophically nonsensical or at least linguistically ambiguous. (There are good philosophical arguments against egoism and it is irrational to ignore them.)

      3. Anonymous

        My utility function has terms in it for the happiness of other people. If rationality means maximizing one’s utility, then it is rational to care about what happens after my death.

        “Rational agents should WIN. If your utility function has a term in it for others, then win their happiness. If your utility function has a term in it for a million years hence, then win the eon.”

    2. gwern

      If that’s important, there’s always sperm cell donation!

      (Yes, I realize American sperm banks are very selective and that as young as he is, Yvain may not pass screening. Fortunately, as he is in a highly-credentialed career with high potential future income, it is very easy for him to move somewhere – or just take a sabbatical for a year or something – like England where apparently Yvain’s sperm will be in huge demand.)

      1. David Gerard

        I’d be amazed if (as a doctor) Scott’s sperm were not in huge demand in the US as well.

        In any case, Scott – I still strongly recommend this. Improve the world and someone else gets the 18 years of work raising the kid.

      2. MugaSofer

        Aren’t there fears of paternity becoming an issue with that? Y’know, cos the legal system be crazy?

        1. gwern

          As far as I know, there is close to zero risk of being sued for child support, if that’s what you mean. There have been a few cases, IIRC, but typically they involved things like the sperm donor visiting or sending presents, which would not be a concern for just straight donators. The main concern if one does the England route is that sperm donors are not kept 100% anonymous: the kid can look you up when he turns 18, IIRC. But other than the risk of some awkward conversations and whatever effort it takes to donate, there seems to be no downsides IMO.

        2. David Gerard

          I can’t speak to the US situation, but I’d be surprised. Do you have a link to anything specific on the subject?

          In the UK, if you donate in an HFEA-approved manner, you have neither parental rights nor responsibilities. The child has the right to ask for your name and your contact details (from a couple of decades before) when they turn 18; you have no right to ask for theirs, but you can ask if your sperm has been used (they don’t tell you, you have to ask them).

          In the UK, if you donate sperm, it’s almost certain to be used, because donations went way down when that bit about the child being able to contact you at 18 came in. As a child who was adopted, I fully concur with the reasons for that rule coming in – the rights of the child produced are the overriding consideration in the gamete donation rules as they stand – but it’s actually a problem (approximately 300 donors a year for 500 donations needed).

Comments are closed.