"Talks a good game about freedom when out of power, but once he’s in – bam! Everyone's enslaved in the human-flourishing mines."

The Dark Side Of Divorce

A while ago I read The Nurture Assumption and found myself convinced by its basic thesis that genetics completely trumped parenting.

The argument was that there are lots of studies showing that parenting has important effects – for example, if parents yell at their kids, their kids will turn out angry and violent, or something. But these studies neglect possible genetic contributions – angry violent parents are more likely to yell at their kids, so maybe the kids are just inheriting genes for anger and violence. A lot of parenting studies are subject to these kinds of confounds. And one of the best tools we have for disentangling them – behavioral genetics twin studies – very consistently show that most important outcomes are 50% genetically determined, 50% determined by “non-shared environment”, and almost completely unrelated to the “shared environment” of parenting. Therefore, we should conclude that pretty much all of the effect supposedly due to parenting is in fact due to genetics, and it doesn’t matter much what kind of “parenting style” you use unless it can somehow change your child’s DNA.

One of the stories I most remember from the book – and I’m sorry, I don’t have a copy with me, so I’m going from memory – was about the large literature of studies showing that children of divorce raised by single mothers have worse outcomes than children of intact two-parent families. This seems like a convincing argument that children need both parents to develop properly, which if true would be a shared environmental effect and an example of why good stable parenting is necessary.

But other studies found that children who lost a father in (for example) a car accident had outcomes that looked more like those of children from stable two-parent families than like those of children of divorce. So maybe the divorce effect doesn’t reflect the stabilizing influence of two parents in a kid’s life. Maybe it reflects that the sort of genes that make parents unable to hold a marriage together have some bad effects on their kids as well.

(damn you, Rs7632287! This is all your fault!)

It’s compelling, it’s believable, and I believed it. Unfortunately, I recently had the time to double-check, and it doesn’t seem to be true at all.

The best introduction to divorce research I could find was Amato & Keith’s meta-analysis Parental Divorce And The Well-Being Of Children. It looks through 92 studies that compare children of divorced and non-divorced families and finds that “children of divorce scored lower than children in intact families across a variety of outcomes, with the median effect size being 0.14 of a standard deviation,” this last clause of which is almost New Cuyaman in its agglomerativeness.

This is a small effect size, and indeed most of the studies they’re looking at aren’t even significant. But once agglomerated together they become very significant, and the analysis tries to determine the cause. The most popular proposed causes are “children in divorced families lose the benefits of having two parents”, “children in divorced families are in economic trouble”, and “children in divorced families have to deal with stressful family conflict.”

Although there’s a little bit of evidence for all three, in general the evidence lines up for the last one of these – the family conflict hypothesis.

If the problem is not enough parents or not enough money, then having the custodial parent (usually a single mom) remarry ought to help a lot, especially if she marries somebody wealthy. But usually this doesn’t help very much at all.

If the problem is not enough money, then children of divorce should do no worse than children of poor two-parent families. But in fact they do, and children of divorce still do worse when controlled for income.

If the problem is not enough parents or not enough money, then these ought to persist over time if the custodial parent doesn’t remarry or get richer. But if the problem is stressful conflict, then it ought to get better over time, since the stress and conflict of the divorce gradually becomes more and more remote. Although there are some dueling studies here, the best studies seem to find the latter pattern – bad outcomes of divorce gradually decrease over time.

If the problem is stressful conflict, then children of divorce ought to do no worse than children in families full of stressful conflict who are nevertheless staying together. Indeed, controlling for the amount of stressful conflict within a family gets rid of most of the negative effect of divorce.

Therefore, although there was some evidence for all three hypotheses, the stressful conflict hypothesis was best-supported. But the stressful conflict hypothesis could also explain the pattern where kids whose fathers died in car accidents don’t show the same pattern of problems as children of divorce. Having a parent die in an accident is no doubt traumatic, but it’s a very different kind of trauma from constant familial yelling and bickering.

More to the point, the genetic explanation of divorce has been investigated specifically in at least four studies that I know of, using different methodology each time.

Brodzinsky, Hitt, and Smith studied the effect of divorce on biological versus adopted children. They were unable to find any differences in the level of disruption and poor outcomes.

O’Connor, Caspi, DeFries, and Plomin (yes, that Plomin) also studied biological versus adopted children. They found that biological children showed a stronger effect on academic achievement and social adjustment (consistent with genetic explanations), but adopted children showed an equal effect on behavioral problems and substance use (consistent with environmental explanations).

Burt, Barnes, McGue, and Iacono use a different methodology and compare children whose parents divorced when they were alive with children whose parents divorced before they were born. Presumably, only the former group get any environmental stress from the divorce, but both groups suffer from any genetic issues that caused their parents to split. They find that the negative effects of divorce are mostly limited to the group whose parents got divorced when they were alive, consistent with an environmental explanation.

Finally, a bunch of people including Eric Turkheimer get the requisite twin study in and compare the children of pairs of identical twins where one of them got divorced and the other didn’t (where do they find these people?) Somehow they scraped together a sample size of 2,554 people, and they found that even among children of identical twins, the children of the divorced twin did worse than the children of the non-divorced twin to a degree consistent with the negative effects not being genetic. They tried to adjust for characteristics of the twins’ spouses, but that’s the obvious confound here. I look forward to seeing if future researchers can get a sample of pairs of identical male twins who married pairs of identical female twins, one couple among whom got divorced.

So I owe mainstream psychology an apology here. I was pretty sure they had just completely dropped the ball on this one and were foolishly assuming everything had to be social and nothing could be genetic. In fact, they were only doing that up until about ten or twenty years ago, after which point they figured it out and performed a lot of studies, all of which supported their idea of the stress of divorce having significant (though small!) non-gene-related effects.

And although I haven’t had time to look through them properly yet, here’s a study claiming that the association between fathers’ and childrens’ emotional and behavioral problems is “largely shared environmental in origin”. And here’s a study claiming that “analyses revealed that [shared environment] accounted for 10%-19% of the variance within conduct disorder, oppositional defiant disorder, anxiety, depression, and broad internalizing and externalizing disorders, regardless of their operationalization. When age, informant, and sex effects were considered, [shared environment] generally ranged from 10%-30% of the variance.”

So the shared environment folks haven’t completely dropped the ball, some of them seem to be fighting back, and it will be interesting to see where this goes and whether anybody is able to reconcile the different evidence.

One likely talking point: shared environment and childhood situation obviously impacts things during childhood. For example, if you have parents who are mean and abusive, this can make you stressed and you don’t get enough sleep and then maybe you do really badly at school. But once you get out of that environment, your academic abilities will revert to whatever your genes say they should be. The Nurture Assumption never denies this and is absolutely willing to admit that shared environment can affect outcomes during childhood, although even there less than one might expect. This also seems to be the tack Plomin is taking when he discusses the Burt study.

But studies have found that the negative effects of divorce can last well into adulthood. On the other hand, none of those studies have been the ones that compare genetic and environmental effects, and I get the feeling their quality is kind of weak. So it’s not completely ruled out by the data that the short-term effects of divorce are robust and environmental, but the long-term effects of divorce are spurious and/or genetic. But this seems kind of like fighting a rearguard action against the evidence.

Finally, a sanity check. Suppose your parents get divorced when you’re 16. Your high school grades drop and your behavior gets worse. Maybe you fail a couple of classes and start using drugs. The couple of classes failed mean you’re going to a second-tier instead of a first-tier college, and the drug use means you’re addicted. How does that not affect your life outcomes, even if five years later you’ve forgotten all about whatever psychological stresses you once had?

Overall I am less confident than before that shared environment is harmless.

And while I’m bashing Nurture Assumption, I don’t remember the exact arguments used against birth order effects, but we found such impressive numbers on the last Less Wrong survey that I’m not very impressed with the claims that they don’t exist.

This entry was posted in Uncategorized and tagged . Bookmark the permalink.

205 Responses to The Dark Side Of Divorce

  1. Samuel Skinner says:

    ” In fact, they were only doing that up until about ten or twenty years ago, after which point they figured it out and performed a lot of studies, all of which supported their idea of the stress of divorce having significant (though small!) non-gene-related effects.”

    I should note The Nurture Assumption was originally published in 1999; the book can be both accurate and wrong due to that.

    “And while I’m bashing Nurture Assumption, I don’t remember the exact arguments used against birth order effects, but we found such impressive numbers on the last Less Wrong survey that I’m not very impressed with the claims that they don’t exist.”

    She was responding to the rather dramatic claim that birth order was a predictor for the protestant reformation and scientific revolutions (and yes this wasn’t a strawman as there was a book with that thesis that had come out).

    • Scott Alexander says:

      I remember that, but I also think I remember denying that birth order effects existed at all.

      • Steve Sailer says:

        Here’s my old review of Judith Rich Harris’ book in National Review:

        http://www.isteve.com/nurture.htm

        I was generally positive, but:

        To show that peers outweigh parents, she repeatedly cites Darwinian linguist Pinker’s work on how young immigrant kids automatically develop the accents of their playmates, not their parents. True, but there’s more to life than language. Not until p. 191 does she admit — in a footnote — that immigrant parents do pass down home-based aspects of their culture like cuisine, since kids don’t learn to cook from their friends. (How about attitudes toward housekeeping, charity, courtesy, wife-beating, and child-rearing itself?) Not until p. 330 does she recall something else where peers don’t much matter: religion! Worse, she never notices what Thomas Sowell has voluminously documented in his accounts of ethnic economic specialization. It’s parents and relatives who pass on both specific occupations (e.g., Italians and marble-cutting or Cambodians and donut-making) and general attitudes toward hard work, thrift, and entrepreneurship.

        Nor can peers account for social change among young children, such as the current switch from football to soccer, since preteen peer groups are intensely conservative. (Some playground games have been passed down since Roman times). Even more so, the trend toward having little girls play soccer and other cootie-infested boys sports did not, rest assured, originate among peer groups of little girls. That was primarily their dads’ idea, especially sports-crazed dads without sons.

        While millions of parents sweat and save to get their kids into neighborhoods and schools offering better peer groups, Mrs. Harris redefines this merely as an “indirect” parental influence. She claims modern studies can’t find predictable relationships between “direct” influences (i.e., different child-rearing styles) and how children turn out. But that may be merely an inherent shortcoming of these non-experimental analyses. For example, she asserts (not necessarily reliably) that studies prove it doesn’t matter whether mothers work or not. But the same methodology would report that it doesn’t matter whether you buy a minivan or a Miata, since purchasers of different classes of vehicles report roughly similar satisfaction. In reality, women don’t randomly choose home or work; they agonize over balancing career and family. They tailor their family size to fit their career ambitions and vice-versa. Mothers then readjust as necessary to best meet their particular families’ conflicting needs for money and mothering. For instance, a working mother might quit when her second baby proves unexpectedly colicky, then return when the children enter school, then shift to part time after her husband gets a big raise. That’s bad for these studies, but good for their kids.

        Finally, why do mothers care so much? Disappointingly for a Darwinian, Mrs. Harris blames it on The Media.

  2. Izaak Weiss says:

    Maybe I’m misunderstanding the final paragraph of your post, but wouldn’t birth order effects imply that having older/younger siblings affects you, which is a nurture hypothesis? Or is there something weird going on with genetics that I don’t understand?

    • Scott Alexander says:

      Yes, I think the book’s denial of birth order effects was part of its general denial of nurture hypotheses.

      • Ben Southwood says:

        I remember Sarah saying somewhere that birth order effects are just an artefact of different size families. But she would have to give the full details as I can’t remember them exactly.

      • Lee Kelly says:

        Not true. Harris also proposed her own nurture hypothesis. Her argument, rather, was that most of the usual suspects trotted out by social scientists to explain why people are different don’t square with the evidence. At best, parenting styles only account for a very small amount of the variance of personality traits (10% if we’re being generous). Most of the variance appears to be a combination of genes and random noise, and the rest is up for grabs.

    • suntzuanime says:

      There are plenty of nature things that could be the result of birth order (the later you are in birth order the older the womb/sperm you come from, and it’s possible that giving birth/raising a child would affect the nature of the womb beyond just the effects of age). I’m not really up on the literature on birth order so they might have ruled this out, but it’s not necessarily a matter of nurture.

      • Anonymous says:

        the prenatal hormonal environment is supposed to be affected by birth order

        • Richard Gadsden says:

          There are also supposedly some effects that are particular to the mother having had a son prior to your birth, so we should probably separate out not just birth order but:

          * Do you have an older brother?
          * Is your closest older sibling a brother?

          • haishan says:

            In particular, there is a good amount of evidence that fraternal birth order is associated with homosexuality.

          • Vaniver says:

            In particular, there is a good amount of evidence that fraternal birth order is associated with homosexuality.

            The last I heard, this data was better explained by the ‘male-loving gene’ hypothesis, where a common cause increases both fertility in women, same-sex attraction in men, and (I think?) percentage of children born male, but I think the sample sizes here are small enough that distinguishing between hypotheses is difficult.

          • RCF says:

            @Vaniver

            Shouldn’t that mean that having a lot of younger brothers would be just as correlated with homosexuality as having a lot of older brothers?

  3. gattsuru says:

    Small sample size, but from anecdote, being a child of divorce doesn’t seem to be a single event. The procedures of the divorce itself are really traumatic to the parents in even the best of situations, but the stressors applied to the children often may not feel as emotionally significant for years. Dealing with a guardian ad litem or living from a duffel bag shuffled from house to house is costly, but emotionally it’s built to be something people can contextualize. Ten years later, trying to weight who can be invited to graduation or a wedding without needing armed security isn’t costly in the classical sense, but is much more emotionally draining.

    Not everyone experiences this — a good many cut off part of the family entirely — but it seems the more common response, and something that shows up at least a few times per year. Three or four years after a divorce is finalized, you still have to deal with reviews of custody, and ten years after, both biological parents may want to get time during the Holidays.

    This may be a matter where the hedonic treadmill doesn’t apply: like long-term unemployment or serious disability, it’s something glaring at you pretty often in a way that seems to violate ‘normality’ in a way most environmental stessors don’t.

    ((I don’t include my own experience here, because I’m really not a typical example, and because the divorce coincided with a chronic disease flareup that pretty much overwhelmed everything else.))

    On the other hand, the monetary issues may also be more present than you’d expect: remarriage after divorce face a much more limited dating pool with more fellow divorcees. and groups of divorcees seem surprisingly prone to splitting into double-income no-kids families or multiple-step-sibling families despite environmental pressures making male primary custody rarer.

    • Anonymous says:

      might the flareup be caused by the divorce? but my real question is: why the double parens?

      • The Anonymouse says:

        Obviously, because the post tells us that two parens are better than one.

      • gattsuru says:

        While stress does act as a trigger for many chronic illnesses, in this case the chronic disease flareup predated awareness of serious parental conflict by nearly half a year. It’s possible that I subconsciously picked up on a lot of small cues pointing to such conflict in a way that mattered, but those were apparently in place for quite a long period beforehand that wasn’t stressful.

        And… I wasn’t exactly the most socially perceptive child.

        The double-parens is just to set asides that a disjoint in purpose from the central post, but apparently the Teleos is to enable The Anonymouse’s amazing bad pun.

    • Deiseach says:

      Dealing with social housing clients, we’re not dealing in general with people who are very well-off (we’re seeing a recent up-tick in people applying for housing/claiming homelessness due to the Irish economic meltdown with the bursting of the property bubble and the banks crashing, since houses are being repossessed due to defaults on mortgages – ironically, since there’s signs of a recovery in the property market, it is now worth the banks’ time to seize property for sale).

      Marriage breakdown ending in separation/divorce (or cohabiting couples splitting up) is traumatic for everyone, from what I can see.

      Dealing with a guardian ad litem or living from a duffel bag shuffled from house to house is costly, but emotionally it’s built to be something people can contextualize.

      I wonder if it is something children can emotionally contextualise; one custody agreement I saw had “Child A will be picked up after school on Monday, stay overnight with father, go to school Tuesday morning and return to mother; Child B will then be picked up after school on Tuesday by father, stay overnight, go to school Tuesday morning, return to mother; Child A will then…/Child B will then…”

      That kind of ‘whisked in and out the door, hardly time to change the sheets on the bed, who’s collecting me today and where am I going’ upheaval must have some kind of effect. Add in that the parents, as I said above, aren’t well-off and so it’s hard for them to be able to juggle schedules, they can’t pay childminders or have the kids in daycare etc., and there is constant sniping about “you owe maintenance/I lost my job, I can’t pay my own rent let alone maintenance/well, without maintenance we’re homeless” and that level of constant everyday stress has to be wearing.

      Then either parent (or both) takes up with new partner. Usually it’s the father, usually there’s a new baby with the new partner on the way (I have no idea if this is the men proving their virility and attractiveness after their marriage failed or what) and then there’s no longer the room/money/time for the kids of the first marriage. This applies even more when the kids are in the custody of the mother, who takes up with new partner and produces new baby which takes up all time/attention/money in the home.

      Or the fathers have no contact at all, either because they take off for parts unknown, or they don’t get custody. It is very unusual to see a father with custody of the child/children, it only happens in very dire circumstances, and given we’ve had a social worker who considers a woman to be a good mother because when she shoots up, she always turns her back so the child can’t see, you can guess the circumstances have to be really dire to lose the mother custody.

      Also, there tends to be an awful lot of moving around. Parents or cohabiting parnters break up. Mother, if she’s an existing tenant, demands transfer because ex-partner (and his new partner) are living in the vicinity. Upheaval and disruption from move, with ensuing uncertainty and lack of control, result. Kids get dragged from pillar to post between one parent moving, the other parent moving, etc.

      And then five, eight or ten years down the line, we see the kids in failed relationships of their own, with a kid or two or more, looking for social housing. Besides the dropping out of school and the criminal records (we have a regular habit of checking the court cases in the weekly local paper to see which of our existing tenants/applicants/partners of same are up on charges and likely to be going to prison) which do arise from the falling between two stools when it comes to parental raising, the uncertainty and stresses during their childhoods at a vulnerable time.

      So whatever about genetics for shiftlessness and impulsiveness, I can attest to anecdotal evidence that environmental stresses have a big part on top of everything else.

      • nydwracu says:

        I wonder if it is something children can emotionally contextualise; one custody agreement I saw had “Child A will be picked up after school on Monday, stay overnight with father, go to school Tuesday morning and return to mother; Child B will then be picked up after school on Tuesday by father, stay overnight, go to school Tuesday morning, return to mother; Child A will then…/Child B will then…”

        That kind of ‘whisked in and out the door, hardly time to change the sheets on the bed, who’s collecting me today and where am I going’ upheaval must have some kind of effect.

        It does. Total mess. Fucked up my grades too — it was hard to make sure all my textbooks were in the right place every day.

  4. AR+ says:

    But these studies neglect possible genetic contributions – angry violent parents are more likely to yell at their kids, so maybe the kids are just inheriting genes for anger and violence.

    I have wondered* if a sufficient degree of genetic determinism and some forms of timeless ethics would lead to the conclusion that most abused children have the parents they deserve.

    *This is not an endorsement of said conclusion.

    • The decision algorithm deserves what it gets, but the child doesn’t deserve to be stuck with that decision algorithm.

      • How is it meaningful to separate the “decision algorithm” from the child? Calling it a “decision algorithm” is obfuscating the fact that the “algorithm” is none other than the person who makes the choices and carries them out.

        I’m definitely not endorsing the OP’s theory, but this evasion is seriously bizarre.

        • Deiseach says:

          You’re saying the child chooses the parent? The parents are the ones who choose to have the child; if they’re angry/violent they may pass on those genes and will provide an angry/violent environment, but I really don’t see how the child can be said to “deserve” their genes/environment from a choice they had absolutlely no input into, or control over.

          • I think that you meant to address this to the OP, not to me. But in any case, the hypothesis seems to be that the child retroactively “deserves” abuse for their later adult misbehavior. The backwards-in-time causation aspect of this is what is implied by the word “timeless”.

          • Deiseach says:

            But we are living in time, and so any possible future crimes or offences we commit are yet to come (and may not come at all), so how can today’s neglect or abuse be “deserved” by someone who in fifteen years’ time may do something bad or wrong?

            I mean, I can certainly forecast in my job that the children in current messed-up situation are themselves, five or eight years down the line, going to replicate the same situation and have kids themselves who are going to be dragged around in the same messed-up way they are currently being raised, but I certainly would not and cannot say “Because I can prognosticate with some success, because I have already had evidential proof of similar situations that I have witnessed where Child X at the school where I worked then went on to be Teenager X at the troubled teen centre where I worked and is now Young Adult X with a few convictions for petty crime, a couple of kids by different fathers/mothers, and is now unemployed and looking for social housing at the local government department where I work, that this child will follow the same path in life – absent drastic intervention – and because they will IN FUTURE be a shiftless wastrel, they deserve RIGHT NOW the suffering they are undergoing.”

            No. Unless you’re going to measure with the same measure for everyone, and OP deserves tomorrow to be mugged and beaten up in an alleyway and put in hospital with internal injuries because, looking at it timelessly and retrospectively, they are going to do something not-nice in five years’ time.

          • But we are living in time… so how can today’s neglect or abuse be “deserved” by someone who in fifteen years’ time may do something bad or wrong?

            Well, exactly. This is why most people don’t consider timeless ethics to be a workable system for actual humans.

        • social justice warlock says:

          Hugging the premise here, this seems analogous to a case in which someone is congenitally irresponsible, or engages in compulsive self-harm. We can hold that they are the cause of their misery without holding that they deserve it – on the hereditarian assumptions of the thought experiment, euthanasia of the algorithm via eugenics is the obviously compassionate choice.

    • Illuminati Initiate says:

      I find the idea of anyone deserving bad things happen to them to be kind of abhorrent.

      • call_me_aka says:

        Do you find the idea of anyone deserving good things happen to them to be abhorrent?

        • I believe that pain is bad. There are all kind of complications – sometimes not inflicting pain (e.g. prison) may be worse, and if Alice does something particularly dangerous I’d prefer any pain to fall on her rather than on Bob (ceteris paribus) – but it’s perfectly possible to prefer a lower level of pain to a higher level for any one person.

          • call_me_aka says:

            Sure, but do you prefer a higher level of pleasure to a lower level for any one person?

            I ask not because I’m trying to make a case for punishment, but because I suspect that the valence of any given statement about desert makes a difference in how we respond to it. I too find the idea of anyone deserving bad things to happen to them abhorrent, but the idea that some people might deserve better things than others makes me somewhat less angry, and the idea that people might deserve good things annoys me in the abstract but doesn’t activate any moral instincts. Which makes roughly zero sense if all you’re doing is rejecting moral desert.

          • Protagoras says:

            Psychologically, people react differently to reward than to punishment, despite the lack of clear distinctions between rewards and withheld punishments or between punishments and withheld rewards in many practical situations. Specifically, punishment seems to have all sorts of bad side effects, which are not shared by rewards. So it may be a beneficial attitude to have to be more comfortable with rewards for the deserving than with punishments for the undeserving, even in the abstract (though the lines get even murkier with greater abstraction).

      • Wirehead Wannabe says:

        Seconded. Punishment before doing something bad does nothing to correct said behavior. Wanting bad people to suffer is a lost purpose.

        • Ghatanathoah says:

          I used to think this, but now I’m not so sure. Yes, punishing bad people does have the effect of deterring bad behavior. That’s probably why we evolved to value that. But that doesn’t mean that that’s why we should value it. We evolved to value happiness because engaging in happiness-causing behaviors increased our reproductive fitness, but that doesn’t mean that seeking out happiness when it doesn’t increase reproductive fitness is a Lost Purpose.

          I’m concerned that maybe the rejection of deserts and punishment is an example of Hollywood Rationality rather than Real Rationality. Taking a behavior that we value in and of itself, but that also has beneficial long-term consequences, and then saying that the beneficial long-term consequences are all that matter, is textbook Hollywood Rationality.

          • jaimeastorga2000 says:

            Thank you!

            Yes, if we can value non-reproductive sex with a loved spouse, and if we can value eating fatty meat and sweet fruits in an environment of caloric abundance, then we damn well can terminally value punishment and vengeance for their own sake.

          • Jadagul says:

            Jaimeastorga2000: We certainly can terminally value punishment and vengeance for their own sake. Most people do. But it seems pretty horrifying.

          • Army1987 says:

            @jaimeastorga2000

            Can != must. The idea of deliberately hurting someone revulses me, no matter how wicked they are, and the only reason I’m OK with the police etc. doing so is that I figure that the positive instrumental value outweighs the negative terminal value. And I wouldn’t want to self-modify to change this.

          • Illuminati Initiate says:

            I don’t value punishing people in of itself though. You might, but I don’t, this is not a matter of rationality but of different preferences (all rationality is instrumental, there is no such thing as an irrational or rational preference)

            Thinking about this though, I do seem to think greater sacrifices can be made of people to stop bad things from happening when they are doing those things intentionally. For example, lets say two people are driving. They will each hit and permanently cripple someone. For Driver A it is an accident. Driver B is maliciously trying to commit vehicular homicide, but will not go on to kill others and a superintelligence is about to immanentize the eschaton/kill everything so deterrence does not matter anymore. I can intervene by killing the driver. I would choose to kill Driver B, but not Driver A.

            This does not mean I value punishment in of itself though. I still only support that as a deterrent, for example I find the idea of Hell (or even temporary purgatory-type afterlife punishment) for anyone horrible (It makes no sense as a deterrent. Even if I accepted that God was right in making the universe so shitty that people can even do such horrible things in the first place (which I don’t), He could just lie about Hell if He wanted to use it as a deterrent).

          • Anonymous says:

            Talk is cheap. I don’t think any of you are introspective enough or honest enough for me to trust your claims about what you value.

          • Ghatanathoah says:

            @Illuminati Initiate
            I’m not sure myself whether I value punishment as an end in itself. It’s just that I recently realized that I dismissed doing so as “irrational” a long time ago and need to reevaluate that. I realized that most people value punishment in consequentalist and agent-neutral terms (by agent neutral I mean that people seem to think that if a bad person suffers due to bad luck that that “counts” as a punishment the same way that a punishment purposefully inflicted by another person would).

            I completely agree with your horror of the idea of hell and purgatory, however. I may not be sure if I value punishment or not. But I’m pretty darn sure that I disvalue disproportionate punishment. There is simply no possible crime that an eternity in hell would not be a disproportionate punishment for.

            Plus, valuing punishment without regards to proportion leads to the conclusion that a world full of bad people being tortured forever is better than a world of good people living happy, but finite lives. This seems obviously insane to me.

          • Illuminati Initiate says:

            @anonymous

            Define value then. I might occasionally have avengeance fantasies or something, but my moral preferences outweigh any desire for such things to actually happen.

            @Ghatanathoah

            Well, part of the reason I brought up Purgatory in addition to Hell was to show that it was not just about disproportionality. Punishment in the afterlife is invisible and so is pointless as a deterrent (because lying about it would work just as well). (And of course an all-powerful God could easily stop all bad things from happening if He wanted to).

          • jaimeastorga2000 says:

            @Jadagul: It seems horrifying from your perspective, because you do not have those values. To me, not valuing vengeance and punishment seems pathetic.

            @Army1987: So, to some obvious limiting cases… you do not feel that the perpetrators of the Hi-Fi murders deserved to die? And if some man committed a similar atrocity against your loved ones, you would not value hurting or killing him?

            @Illuminati Initiate: Agreed. Terminal values are terminal values. The utility function is not up for grabs.

          • Matthew says:

            @Jaime

            And if some man committed a similar atrocity against your loved ones, you would not value hurting or killing him?

            I think this suggests an odd use of the term “terminal values.” Pretty much everybody has a desire to punish built into their system 1, as spillover from something that is adaptive for game-theoretic reasons. But if someone is consciously choosing to override that system 1 impulse, wouldn’t you agree that their explicit, system 2 objection is their terminal value?

          • jaimeastorga2000 says:

            @Matthew: I am asking whether he would reflexively and consciously agree with the inborn instinct to kill the hypothetical killer, not whether he would have the instinctive response in the first place (which, as you point out, most everyone does).

          • Jadagul says:

            Jaimeastorga2000: Yeah. You’ll note I didn’t attempt to mount an argument, because terminal values don’t really admit to arguments. Sometimes I just feel the need to register my values and assert that people with those values do in fact exist.

          • Jaskologist says:

            Let us imagine a different Breaking Bad ending:

            Walter White: I’ve decided to stop making meth. Also, I am sorry for all those people I killed.
            Addicts: We’re not addicted anymore, so nobody will be incentivized in the future to make meth!
            The Law: Well, I guess you’re good to go then, Walter. Enjoy all the moneys!

            If you find this unsatisfying, congratulations, you probably do value punishment after all.

            (Feel free to write a similar ending for Dexter or some other “the bad guy is the main character” show. There’s a good chance you’ll even beat the actual Dexter finale.)

          • Jaskologist says:

            We evolved to value happiness because engaging in happiness-causing behaviors increased our reproductive fitness, but that doesn’t mean that seeking out happiness when it doesn’t increase reproductive fitness is a Lost Purpose.

            Really, I think Ghatanathoah won this thread before it began right here. Stop picking arbitrary values and pretending they should be universalized. You’re like cells optimizing themselves to take in as much insulin as possible. If you’re going to do that, at least be honest about it.

          • Jadagul says:

            Jaskologist: I haven’t watched Breaking Bad, but…like, if the story finds a plausible way for everyone to get a happy ending, that’s a good thing. Or even an implausible one if the story is surrealist enough that you don’t mind.

            We don’t know how to make it so that no one suffers, ever. But that’s a _problem_.

          • Illuminati Initiate says:

            What people value in their fiction is not the same as what they value in real life. There is a reason dystopian stories are far more common than utopian ones.

            Edit: Also, I’ve never seen Breaking Bad, but isn’t Walter supposed to be thought of as a villain protagonist?

      • Deiseach says:

        It depends on what the person does/has done, the amount of freedom and knowledge they had, and the harm they have done to others, and what the punishment/bad things are.

        If I continue (say) to drink so heavily that my liver collapses, in a sense I ‘deserve’ this because this was predictable harm I was warned about that I had the opportunity to prevent. In another sense, of course, nobody “deserves” liver failure, no matter how bad their choices or how bad a person they are.

  5. The Anonymouse says:

    While certainly divorce is disruptive, it is dangerous to just assume that all divorces lead to continuing bitter conflicts between the parents. I know of several examples where, after a period of unavoidable disruption, the coparents then spend the next years and years getting along famously.

    Is this as good as an intact family with two happy parents? Probably not. But it’s much better than the bitterness and fighting that most people (and this post) assume when they hear “divorce.”

    • Kaj Sotala says:

      My parents divorced when I was quite young. I didn’t even realize that they were divorced until my mom happened to mention it many years later. They did live separately with my dad occasionally coming to visit to spend time with us, and at one point I visited him on weekends, but I didn’t realize that there was anything unusual with that, or that there was anything “wrong” with the relations between my parents.

  6. creative username #1138 says:

    Finally, a bunch of people including Eric Turkheimer get the requisite twin study in and compare the children of pairs of identical twins where one of them got divorced and the other didn’t (where do they find these people?)

    Don’t know about psychologists but economists really love twin festivals for their studies. Talk about a self-selected group of twins.

    • Paul Torek says:

      According to my wife the psychologist, another source of high N for twin studies is fast-and-loose definitions. See “identical twins raised apart” in studies of the genetics of schizophrenia. The effect sizes shrank with successive studies, in part because of a tightening of definitions. Gwern discussed the shrinking effect sizes somewhere, as part of a larger study of publication bias, but I couldn’t find that post, alas.

      • Douglas Knight says:

        That sounds backwards to me. If you want to know the role of genetics, you compare identical twins to fraternal twins. The purpose of adoption studies is to measure the role of parenting.

  7. ishaan says:

    Did the studies seem to indicate that the “stressful conflict” centered around the divorce itself, or the circumstances that led to the divorce?

    In other words, practically speaking do people think this means we should be discouraging divorce so as to avoid “stressful conflict”? – or should we be making it easier and faster so that parents undergoing “stressful conflict” don’t continue engaging in it for prolonged periods of time?

    • lmm says:

      I don’t have studies to hand but my memory is that when controlled as far as possible, couples in the same circumstances and at the same level of conflict generally become much more conflicty if they divorce than if they don’t.

      • Anonymous says:

        That makes sense, and multiple people who’ve replied to me have said now some version of this … but I just did a second reading, and how does that account for Burt et al? (Recap: divorce before birth wasn’t as bad as divorce after birth)

        Edit: Oh, this is still ishaan, forgot to fill the name field

    • Jaskologist says:

      This was the original justification for no-fault divorce. How’s that working out so far?

      • social justice warlock says:

        Marital satisfaction is higher than ever, IIRC. Parental satisfaction may be as well, though I may just be confusing that with time-spent-on-children measures. Of course there are probably costs borne as well, and indeed the costs might outweigh the benefits – I’m not in a position to say – but it doesn’t seem that we live in a world where this is sufficiently obvious to merit your use of the rhetorical question.

        I also suspect that without no-fault divorce, but with all the other cultural changes that have happened, people would be more reluctant to form marriages. Of course the resulting marginal cohabitations might end up being mostly stable anyway – or not. Counterfactuals are hard!

        • Lesser Bull says:

          People already are more reluctant for form marriages. At least anecdotally, part of it is the instability of marriage (people don’t want to get married and have kids just to get divorced).

        • Jaskologist says:

          Bragging about how high the marital satisfaction rate has always seemed to me like bragging about how the low the “horses with broken legs” rate is.

          Also, I’m having a devil of a time finding actual figures on marital satisfaction over time. Everything I get is about an individual couple’s lifetime, not broader, country-wide trends.

    • Deiseach says:

      You can make the process of getting a divorce easier and faster, but if people are arguing and bitter while the marriage is breaking down, they’re likely to still be angry and bitter afterwards, and the conflict over custody, maintenance, property division, whose fault it is etc. will rumble on.

      If I want a divorce because I’m sick of the sight of your face and will commit murder if I have to stay with you one week longer, the immediate relief of ending the marriage is one thing, but if I’m then compelled to have some contact with you re: children, even if it’s only paying maintenance out of my salary every month, that is still going to be a sore point because I hated your guts before and paying your lazy ugly ass my hard-earned cash isn’t making me any fonder of you.

      If I do have the kids for access, then I’ll be likely to badmouth you to them, try to bribe them to win their affection from you, and quiz them on any new partners etc. in your life so I can go back to court and have the maintenance payments decreased or stopped.

      • houseboatonstyx says:

        Even a worst-case post-divorce feud at a distance, will not be like living in the same house with two parents fighting in front of you.

        • Morgenstern says:

          You’d be surprised. Not having to live with someone means there’s no reason to hold back anymore, and having to compete over kids / property / money rewards the one who fights dirtiest.

          Obviously this is just one data point but parents can get a lot more vicious once they separate.

          • houseboatonstyx says:

            But “not in front of the children, dear” becomes much much easier. Presumably in zis own bed at home at night, there will not be a fight in the next room. Etc.

          • nydwracu says:

            But “not in front of the children, dear” becomes much much easier.

            Hahahahhaha.

            My parents got a lot nastier after they divorced.

          • Mary says:

            You assume that its ease is a point with the parents.

            They might fight all the more bitterly because the chance is fleeting.

        • Deiseach says:

          Still doesn’t prevent things like (pulling recent examples out of memory) mother preventing father’s name from being put on the birth certificate of new-born and preventing father from coming to the christening.

          You’re going to tell me there won’t be lingering bitterness still there once the kid is old enough to start asking questions about “Who’s my father and why haven’t I ever met him?” from the mother, the mother’s family, the father’s family (“You never met your father because your bitch of a mother wouldn’t even let him come to your christening”), etc.?

          And this is Ireland, not America; there are no really large distances people can end up going. Unless the mother in question moves to the opposite end of the country and takes the kid with her, there will still be family connections/neighbours’ gossip in the area they’re living.

          • Jaskologist says:

            There’s also an underlying assumption here that the bad marriage won’t simply be replaced with another bad relationship. The sad truth is that people tend to make the same mistakes over and over, and if you were attracted to a loser before, you’re probably going to be attracted to one again.

            And that’s compounded by the fact that a divorced woman with kids and stretch marks to match has basically no chance of trading up from the type of man she was able to snag when she was young and baggage-free.

          • Ghatanathoah says:

            @Jaskologist

            And that’s compounded by the fact that a divorced woman with kids and stretch marks to match has basically no chance of trading up from the type of man she was able to snag when she was young and baggage-free.

            Her previous man would have to be pretty darn awful, in order to consider dating a man who cares about things like stretch marks to be “trading up.” Someone who cares a lot about those things is a pretty low quality mate.

          • Jaskologist says:

            I wish her all the luck in the world finding a man who doesn’t care about her appearance. But I don’t expect “I’m going to compensate for my added emotional baggage and financial liabilities with reduced attractiveness” to be a very successful strategy.

    • Troy says:

      I do not know what studies on this say, but I think that our general knowledge of human psychology should lead us to expect that in most cases (though not all), divorce will exacerbate conflict. In general, if you part with someone on bad terms, you leave with a bad impression of that person that then gets amplified in your mind over time. By contrast, if you continue relating to that person, then whatever conflict happened that could have but didn’t lead to you cutting off contact with them tends to recede in importance. (There are exceptions, obviously. I knew a couple who were both violent drunkards and needed to get out of each others’ lives so they could stop blaming all their problems on the other person. But I think these are exceptions.)

      Also potentially relevant is the fact that people tend to overestimate how bad some aspect of their life will be a year from now. People probably overestimate how terrible their marriage will be if they, say, stay married to the man who’s cheated on them.

      • Lesser Bull says:

        Anecdotal evidence:

        My spouse and I both had parents who fought a lot growing up. Hers got divorced, mine didn’t. Mine got over the fighting, hers have fought more. Probably other factors at work, but since you can’t make a clean break where children and money are involved, my guess is that divorce makes thing worse at the margins.

        • nydwracu says:

          Probably other factors at work, but since you can’t make a clean break where children and money are involved, my guess is that divorce makes thing worse at the margins.

          The other alternative is to make a clean break and cut off the children.

          The other other alternative is to, as the children get older, gradually make the existing break cleaner until the children hardly ever see you again and expect that you’ll probably drop off the face of the earth entirely within a few years.

          This is an especially harmful option when everyone on the other side of the family is dead except the parent and one very hermitlike uncle.

          From what I’ve heard, my parents’ divorce was a lot less nasty than most.

      • Deiseach says:

        I don’t really believe there’s such a thing as “no-fault” divorce, since it’s a human failing to seek to cast blame on someone. Certainly there probably are a proportion of cases where both parties can part amicably, but I imagine these are either (a) early in a marriage where there are no children yet or (b) much later, once the kids are all grown and settled themselves.

        Nobody likes to think “It was my fault my marriage failed” so it’s natural to blame the other party as being at least as much, or more, to blame. Throw in children, where there can be custody fights, fights over maintenance (and as I said, from my limited observation a lot of the men go on to new partners, new babies) and simple remaining stress from what ended the marriage (if you broke up because you lost your job or had to take a drastic reduction in pay, couldn’t pay your mortgage and lost your house and are now much more limited in your income, for instance) those won’t go away magically with a divorce.

        • Banananon says:

          I don’t think “no-fault” is supposed to be the parties view of the situtation. My understanding is it’s the judgement of the law that neither partner bears legal responsibility. But take this with a grain of salt; I am not a lawyer, and have not witnessed any divorces first-hand, so my understanding may be off here.

          • Anonymous says:

            Sure it is. We legislate facts in this country. In California, it is illegal to blame your spouse for the divorce.

          • Anonymous says:

            The “no-fault” refers to neither party needing to prove wrong-doing on the part of the other in order to obtain a divorce, instead being able to obtain one at will.

  8. I was sceptical about The Nurture Assumption because it was so weak on homosexuality: implying that homosexuality was a function of school size–i.e. more prevalent in larger urban high schools because more models to follow (p.277).

    Part of the general difficulty about assessing effects of parenting is that kids in the same family can experience quite different parenting. Makes it hard to untangle effects.

    • vjl says:

      “I was sceptical about The Nurture Assumption because it was so weak on homosexuality: implying that homosexuality was a function of school size–i.e. more prevalent in larger urban high schools because more models to follow ”

      This pattern would also be consistent with Greg Cochran’s pathogen explanation for homosexuality (http://westhunt.wordpress.com/2012/02/16/depths-of-madness/).

      One that understandably has not gained much ground in the literature, but does no worse a job of fitting the evidence than any other models out there.

      • Charlie says:

        Looks like it does a pretty bad job to me, between animal homosexuality and the broad and somewhat uniform spacetime distribution of teh gay. Are there any big redeeming features?

        • Anonymous says:

          Are you talking about Harris’s theory or Cochran’s theory?

          Both of them are trying to explain why people raised in cities are more likely to be homosexual than those raised on farms. It’s one thing to object to a theory because it predicts something wrong that the author didn’t think of, but if it is motivated by particular evidence, maybe you should worry that you are wrong about facts.

        • vjl says:

          Is obligate homosexuality common in non-humans? I only know of it in Rams, which certainly could be the result of a similar (the same?) pathogen. Hell… there is even an obvious story for patient zero.

          I don’t think there is any strong evidence in favor of the pathogen theory. That said, I never followed it closely so I am not a good advocate. Prevalence rates are in about the right range and increase with population density. That is all I have.

          The appeal to me is in the weakness of other explanations. Nobody has come up with a remotely convincing evolutionary explanation for how homosexuality balances the other side of the fitness ledger.

          • Among penguin colonies in zoos, most homosexuality appears to be situational but a subset of individuals only ever mate with same-sex penguins, even when opposite sex partners are available.

          • Anonymous says:

            Among mammals, only two species exhibit male obligate homosexuality.

          • Lesser Bull says:

            @anonymous,
            are penguins one of those species, or do you contest AcesUnderGlass’s data.

          • Morgenstern says:

            Penguins aren’t mammals.

          • Anonymous says:

            Penguins are a good example, but most bird examples are problematic, which is a reason to restrict to mammals. The problem with birds is that a lot of the life-long homosexual examples are in species that mate for life, so it is hard to distinguish between obligate and facultative homosexuality (but it also means that the fitness cost is the same).

          • Ghatanathoah says:

            Nobody has come up with a remotely convincing evolutionary explanation for how homosexuality balances the other side of the fitness ledger.

            I don’t know about other animals, but for humans, what about the explanation that anti-homosexual violence increases the fitness of homosexual genotypes, because homosexuals enter into heterosexual relationships out of fear of getting murdered or shunned?

            This theory would predict that regions that are more accepting of homosexuals, and have been for a long time, to have lower frequencies of homosexuality than homophobic areas. I have no idea how to even begin to test that, there’d be so many confounding factors.

      • Anonymous says:

        Both theories address the fact that homosexuality is more common in people raised in cities than farms. That doesn’t make them consistent with each other, but consistent with the evidence. On the contrary, different explanations of the same observation are not consistent with each other.

        • Wulfrickson says:

          Isn’t this massively confounded by the fact that attitudes toward homosexuality are more liberal in cities? I don’t see how self-report data (or most sources of data, really) could get around this.

          Anecdata: I attend a college with a pretty liberal-leaning student body. Every year, the student newspaper conducts surveys of arriving freshmen and graduating seniors, and the surveys include questions about sexual orientation. For a recent class, the portion of male students that identified as gay or bisexual increased immensely – I think fivefold or so – from freshman to senior year, which is easy to explain as “gay people are more honest with themselves when they’re in gay-friendly environments” but pretty hard to explain biologically. (There was also an increase in women who identified as lesbian or bisexual, though far less marked. No word on genderqueer or nonbinary people – sorry, Ozy!)

          • Izaak Weiss says:

            What College? Just out of curiosity.

          • Wulfrickson says:

            I’m trying to be circumspect about my IRL identity (though not entirely; “Wulfrickson” is a very obscure pun of sorts on my real name), so I’d rather not say, but it’s in a city in the northeast US and well ranked academically.

          • Anonymous says:

            That’s not relevant to vji’s comment or my reply.

            Confounding or causing? Are you proposing something really that different from Harris?

            I am very surprised that the effect was larger in men.

          • Wulfrickson says:

            I’m saying that we should be skeptical of claims that homosexuality is more common in cities that don’t try to control for the effects of social attitudes on self-reporting (which, as my anecdote shows, seem at least possibly quite powerful).

          • Anonymous says:

            There are many possibilities. One possibility is that they changed their survey answer without changing their behavior. But you said originally more honest with themselves, which sounds like Harris’s theory that tolerance made them gay.

            But I want to emphasize that this is a very minor point compared to the very serious logical errors of vji’s original comment, where he appeared to deduce the differential prevalence from Cochran’s theory. This is bad for many reasons. One is that it is important to know which evidence is observed directly and to worry about theories influencing observations. The other is that such a deduction cannot be evidence for a competing theory.

          • Wulfrickson says:

            I would venture that “honest with oneself” (or “out to oneself,” if you prefer) and “honest on self-reported surveys” are very correlated, though of course it’s hard to gather more than anecdotal evidence on this.

        • vjl says:

          I wasn’t going to reply, but since you accused me of “very serious logical errors” in a post below, I suppose I should.

          I apologize if it was not clear to you, but when I wrote “This pattern would also be consistent with Greg Cochran’s pathogen explanation for homosexuality”, “This pattern” referred to the evidence and not the theory.

        • Deiseach says:

          But isn’t that as much to do with population density and anonymity as anything? If you have six million people living in a city, as against one hundred thousand living in a rural area, I’d expect any kind of survey about anything (e.g. ‘what is your favourite brand of wellies?’) to have a much higher chance of getting respondents who say “I really like Le Chameau” in the city, than you would in the country (more likely to go “Arrah, the oul’ pair a’ Dunlops I bought in the co-op shop the time I went in to get the cattle drench and calf nuts do me fine”).

          Everyone in the country knows your, your family’s, and your granny’s* business. It’s a lot easier to say “yeah, I’m gay” when you’re living in a place where not even the person next door knows your name.

          *Seriously, I know who someone’s real grandfather is, as against who they think is their grandfather, due purely to the fact that I’m from the country and when my mother used bring me visiting with her, I’d be listening to the old people talking.

  9. Richard says:

    It’s ages since I read The Nurture Assumption, but I vaguely remember that one thing to take away was that children are not socialised by their parents but by their peers.

    This corresponds well with my own personal and informal twin study where the one of my twins with geek friends is a pretty typical geek while the one with stupid jock friends is a stupid jock despite all parental efforts.

    Is this point made in the book, or should I read it again?

    • Vaniver says:

      This corresponds well with my own personal and informal twin study

      Interesting! Monozygotic or dizygotic?

    • Samuel Skinner says:

      Yep, that was her claim. It doesn’t have to be friends, it is that the people who you interact with the most and who have the most influence on you, not your parents and authority figures (unless you spend the majority of your time with them).

      • This might be the most powerful argument for homeschooling that I’ve ever heard.

        • Quixote says:

          Or live somewhere with good schools. Most cities and even tech clusters in suburbs have technology magnet schools which are good and have high achieving student.

        • Paul Torek says:

          It’s funny how the book is often presented (I haven’t read the book; media presentations may be unfair) as if it proves parenting doesn’t matter. Whereas what it does prove, if anything, is that parents who want to have an effect need to be far more involved than is typical.

          • Douglas Knight says:

            “Parents who want to have an effect” are fools. Decisions should be made by weighing costs and benefits. The news that the cost/benefit ratio is higher than previously believed should result in less action.

          • Paul Torek says:

            Whether an act is foolish depends on the agent’s preference rankings. Also, it’s worth considering that the cost/benefit ratio may be nowhere close to constant across the spectrum of time-investment. Maybe Harris’s next book will be on home schoolers.

    • Mary says:

      Yeah, but what she said is that relationship between adoptees’ personality and the rest and their parents is random.

      If your parents have no influence at all on what friends you have — from geographic constraints if nothing else — I suspect that the reality is that birds of a feather flock together, and it’s the personality that lets them track down an appropriate peer group.

    • Sam says:

      Another point of evidence in favour of this theory: children learn the dialect of their peers, not their parents’ spoken dialect. This is slightly more subtle than just learning the dialect of the surrounding area, because there can be diachronic (generational) change (e.g., the ongoing Iberian dialect levelling, which began after broadcast media became a thing; this may also be happening to British dialects under the influence of Southern English).

  10. Shenpen says:

    I don’t know shit about genetics, but to me it is suspicious to conflate genes with human hardware, so to blame genes every time a social/software cause cannot be found. All kinds of stuff can happen to a fetus in the womb. The mother can be stressed, malnourished, can smoke cigarettes, or get punched in the stomach. So by the time a baby is even born you have environmental causes. “Inborn trait” aka “trait there from birth” does not equal genetic. (Even though genetos in Greek would basically mean “born”. Still, let’s not confuse it.)

    For this reason my position would be either demonstrate a gene for it, or it is environmental, but in a broad sense of environmental, not only you don’t need to be born to be affected by the environment (as above), you don’t even need to be _conceived_ as we cannot rule out that mom or dad eating bad or being stressed or really anything, could affect their sperm and eggs or the functioning of the womb.

    And because environmental in this sense is really a broad fucking sense, let’s find another word for it, please.

    • ckp says:

      We can untangle womb effects by comparing MZ to DZ twins.

      • Wulfrickson says:

        Not entirely, actually! There seems to be some confounding from the fact that monozygotes are more likely to share a chorion than dizygotes. See here (unfortunately, I can only find the abstract, not the actual article).

        • Patri Friedman says:

          This just seems, like epigenetics, like the kind of 2nd or 3rd order effect that people who don’t like the strong 1st order claims cite in order to try to contradict them. These things, while real, are too small to reverse the 1st order effects, yet are cited way beyond their degree of significance as a way to fight against the 1st order effects.

  11. naath says:

    So if the the bad results are caused by the stressful situation and constant low-level aggression and so forth… then this tells us nothing about *divorce* at all, it tells us that “your parents’ relationship is unhappy and stressful” is a negative influence on your outcomes.

    I mean, this is all super interesting from an “I want to know everything about how humans work” perspective; but I feel that the media spin on this research is usually “divorce is bad mmkay, stay married!” but staying together might actually make all the stress worse, so how does that improve outcomes? And the difficulty of actually getting a divorce (both social and legal difficulties) may indeed make it worse? Has anyone actually looked at these questions?

    • Richard says:

      Yes they have.:
      paywalled link

      From the abstract:

      The somewhat paradoxical pattern of findings was as follows: (a) The negative associations between parental divorce and various outcomes were found to be generally very similar in Norway and the United States in spite of the great differences in family policy and welfare benefits for single mothers (at the macro level); and (b) Mediational effects of family economic resources were in both countries most marked for the academic achievement area, and the predictive power of such variables was quite similar, again in spite of the great differences in absolute level of the economic resources available to single-mother families in the two countries. The results cast some doubt on the value of the absolute economic deprivation perspective in explaining the results, and the many Norwegian welfare benefits do not seem to mitigate the association between divorce and negative outcomes for the children involved.

  12. Tracy W says:

    The other complicating factor is that often parents divorcing mean the kids move home and move school, which is disruptive to their peer group formation. If income falls, that might be a move to an area with

    And, if one parent remarries and thus household income goes up, that might mean yet another move to a bigger home and thus more disruption.

  13. aretae says:

    I find the evidence in non-humans small pack mammals to be far more compelling than just the stuff you cited above.

    Elephants are the best known of the examples in which the lack of an adult male elephant in the herd leads to adolescent elephants becoming juvenile delinquents.

    Overall, I’ve referenced poor nutrition and divorce as the two known (anti-Harris) childhood environment effects that persist for the last half-dozen years.

    Though…I’m mildly inclined to see the divorce bit as largely a poverty effect. Divorce is (AFAICT) far and away the worst financial thing that can happen to a family. Stress related to money shoots through the roof, and the number of adults about to absorb the stress/comfort one another drops hard.

    If I were studying the topic, I’d study really rich people (who can afford substantial “help”) who divorce to see if their kids experience the same costs. I’d expect a qualified yes, with the magnitude being smaller.

    • Douglas Knight says:

      Yes, looking at mammal models sounds to me like a good idea. But do you know if anyone has systematically studied many species? Elephants sound cherry-picked. Maybe they are the a good candidate because there are interventions, while primates are back to correlational studies.

      • Jaskologist says:

        It sounds like a bad idea to me. Different animals are different, and what holds for one may not hold for another, even within the same family.

        For example, do we go to chimps or bonobos to learn more about human mating patterns? You’ll get opposite answers depending on which you choose, but both are equally related to humans.

        • Douglas Knight says:

          If I had to choose one, I’d choose baboons. But why should I choose one? I specifically rejected that and ask for systematic study of many species.

          • Jaskologist says:

            Why not just study humans directly at that point?

          • Douglas Knight says:

            Animal models have many advantages, such as intrusive data collection and randomized interventions. But the big advantage is that the scientists are more objective. They are less likely to anthropomorphize the subjects, to have politicized input and output.

          • Matthew says:

            Animal models have many advantages, such as intrusive data collection and randomized interventions.

            This seems like an ethical dilemma to me. The more useful a species is as a comparison for humans, the more that suggests they may possess the characteristics that make active experimentation on them problematic.

  14. Lee Kelly says:

    Harris’s other book was No Two Alike, and really should be read as a companion to The Nurture Assumption. In it, she proposes peer effects as a major environmental determinant of adult personality. That is, it’s your socialisation by your peers, and especially social status when young, which have lasting impacts.

    From your description, none of these studies try to separate whether conflict at home is also correlated with a drop in status among peers. That might be an angle to investigate.

  15. JT says:

    Just a minor article consistency issue, the sort that I’d expect most authors not to care much about, but:

    “Unfortunately, I recently had the time to double-check, and it doesn’t seem to be true at all.”

    “all of which supported their idea of the stress of divorce having significant (though small!) non-gene-related effects.”

  16. Quixote says:

    I’ll note that if you believe that divorce, an environmental effect, has significant impact on outcomes than you believe that environmental effects are able to have significant impacts on outcomes. This in turn should increase your prior that any particular environmental effect is significant. If you previously had a very low prior for environmental effects this should be suprising to you and should make a meaningful increase in your priors for other environmental effects.

  17. JayMan says:

    Scott, you don’t owe an apology to the “shared environment” people. But you do owe an apology to us genetics people. :p

    I think, in general, you are not appreciating how difficult it is to address the genetic confound, because like many who do attempt to address it, you are doing it only slipshod service.

    All of those adoption studies, as you note, only reveal what we already knew: the shared environment can have an effect when you’re young. The shared environment is often > 0 in youth, but peters to 0 in adulthood.

    Of course, that’s not even to mention child-to-parent effects. Why assume causation goes divorce -> rotten child behavior?

    And as evidence for lasting effects of childhood family environment in adulthood, you fall back on the same genetically confounded studies? What gives?

    Children-of-twins studies, like twin-control studies, can only be used to show a negative, not a positive. The obvious confound – when you look at divorce – is the twins’ spouses. The DNA of the other parent obviously matters.

    There’s plenty of evidence (see my post More Behavioral Genetic Facts) that the shared environment is indeed exactly zero for both adulthood and indeed many childhood ill outcomes.

    As for your birth order results, that’s so confounded it’s hard to even mention. Did it not occur to anyone there that birth order is an effect of family size? You need to control for that (among other things) to properly look at birth order.

    Judith Harris also deserves an apology, I think. 🙂

    • Izaak Weiss says:

      “The obvious confound – when you look at divorce – is the twins’ spouses.”

      The data from at least one of the studies is described as being from “pairs of identical twins”, which sounds to me like two pairs of twins married, a la this story from huffington post.

      • Anonymous says:

        No, Scott described it as twins marrying non-twins.

      • AR+ says:

        Interesting. A year ago that would have seemed cute to me. Now, it seems that just knowing that Catholic anti-incest law once counted ties by marriage makes this vaguely discomforting. I mean, a marriage between brother* and sister*? Scandal! Incest! If not in the biological sense, then certainly in the sense that, say, going to graduate school at the same university you got your bachelor’s degree from is sometimes called “incestuous.”

        *-in-law

        • Anonymous says:

          I recently double-checked that Edward Bernays, the founder of public relations, was the nephew of Sigmund Freud, and I discovered that he was the double-nephew.

          Did you know that Medieval canon law banned marriages between sixth cousins?

          • AR+ says:

            Yes, but eventually that proved to be too much. Presumably it was easy enough when everybody was in moderately consanguineous tribes, since you just go a few villages over and few people are related to anybody from your own village, but after a while of intermixing it would be nearly impossible to find somebody who didn’t share a great-great-great-great-great-grandparent. Thus, it got changed to only ban 3rd cousin marriages in 1215, but when the population is already highly exogamous, that’s good enough. Compare to modern Middle Eastern consanguinity, where prohibition on 6th, or maybe even higher, cousin marriage might be needed to get any real mixing.

          • Anonymous says:

            No, it didn’t “eventually” become too difficult. No one ever paid any attention to the rule. When they instituted the third cousin rule, they explicitly said “We’re serious this time and no excuses about not knowing your relatives.”

            Also, the sixth cousin rule was devised by confusing different ways of measuring degrees of relation.

          • AR+ says:

            Hmm, that makes sense, since from what I remember, the original rule was, “the degree of separation used in Roman civil law, but calculated in such a way that effectively doubles it.”

            Edit: Yep, anon is right, as can be seen in the text of the original canon, specifically 50 and 51. Also contains this gem of motivated reasoning:

            “The prohibition also is not in the future to affect marriages beyond the fourth degree of consanguinity and affinity; since in degrees beyond the fourth a prohibition of this kind cannot be generally observed without grave inconvenience. This quaternary number agrees well with the prohibition of corporal wedlock of which the Apostle says that “the wife hath not power of her own body, but the husband; and in like manner the husband also hath not power of his own body, but the wife” (I Cor. 7: 4); because there are four humors in the body, which consists of four elements.”

          • Deiseach says:

            You still have to get a dispensation from the bishop to marry your first cousin (as my aunt had to do when she was getting married).

            And that’s part of the plot (though not explicitly stated as such) of Synge’s The Playboy of the Western World; in her small village, everyone is so inter-related already, Pegeen Mike is going to have to look for a dispensation to get married (that’s part of why the outsider, Christy Mahon, is seen as such a prize by the local women competing for his attention: new genetic material!)

            PEGEEN — [with rather scornful good humour.] — You’re making mighty certain, Shaneen, that I’ll wed you now.

            SHAWN. Aren’t we after making a good bargain, the way we’re only waiting these days on Father Reilly’s dispensation from the bishops, or the Court of Rome.

        • Deiseach says:

          Which was the reason Henry VIII (or rather, Henry VII) got the dispensation in the first place to marry his brother’s widow, and then when he wanted to marry Anne Boleyn, argued that the dispensation was invalid, the reason he had no surviving male children was because his incestuous marriage was being punished by God, and the Pope should issue an annulment.

          Katherine’s arguments that her first marriage had not been consummated, and the contradiction involved in “you had no right to issue a dispensation in the first place/you have the right to annul this marriage now” notwithstanding, we know what happened next.

        • lmm says:

          My understanding is that incest taboos were promoted by the church less out of concern for child welfare and more for the sake of preventing noble families concentrating their power.

    • T. Greer says:

      Jayman,

      I think the Burt and D’Onofrio studies are worth a more in-depth break-down/analysis over at your place. Common complaint of the HBD blogosphere is that psychologists don’t try and control for genes. These two explicitly have. It would be helpful to see a longer explanation for what they did wrong and how you think future studies could do better.

      I’d link to it.

      • JayMan says:

        Maybe, but not necessary, in my opinion. They didn’t report significance of their results, so there’s no finding to critique, I’d say.

        Also, see my latest response to Scott.

    • Scott Alexander says:

      “Of course, that’s not even to mention child-to-parent effects. Why assume causation goes divorce -> rotten child behavior?”

      See, this is the attitude I’m attacking here. Yes, you need to check for whether child -> parent effects explain the effects of divorce. Yes, it’s been done several times and all of the good review articles (including the ones mentioned in this post) are careful to mention that it’s been done several times and the the child -> parent effects have been diligently investigated (and found, but there are still parent -> child effects remaining afterwards). See eg here, although there are lots of others. Studies done by Plomin, Turkheimer, etc. are not going to be making rookie mistakes.

      “And as evidence for lasting effects of childhood family environment in adulthood, you fall back on the same genetically confounded studies? What gives?”

      Having found that the studies which investigate genetic versus environmental effects find that it is mostly environmental, it seems reasonable to assume that other studies finding long-term effects are also mostly environmental. I realize that can’t be proven for sure until the same study finds genetic AND long-term effects, and I am very careful to mention that in the post. But in the absence of any study I know of that has investigated it, I think it’s reasonable to combine the best evidence we have and say it points to the likelihood that the long-term effects work the same way.

      “Children-of-twins studies, like twin-control studies, can only be used to show a negative, not a positive. The obvious confound – when you look at divorce – is the twins’ spouses. The DNA of the other parent obviously matters.”

      Yes, this is another thing I explicitly mention in the post. However, what they found was that there was minimal difference between the size of the divorce effect in monozygotic twins versus dizygotic twins. Although one would not expect the non-genetic effect to be zero, because of the spouses, one would expect it to be more significantly less. Twin studies’ ability to show a negative is perfectly consistent with them showing a negative result for major genetic effects in divorce. If a twin study showed that monozygotic twins and dizygotic twins had the same concordance rate for some disease, we would use that information to decide that the disease was not strongly genetic. Again, give Turkheimer some credit here for not running a completely useless study.

      “As for your birth order results, that’s so confounded it’s hard to even mention. Did it not occur to anyone there that birth order is an effect of family size? You need to control for that (among other things) to properly look at birth order.”

      Yes, obviously that is a confound, that is why we specifically set up that analysis after reading Nurture Effect to avoid it.

      I challenge you to come up with any specific way family size could have confounded the LW survey results. The entire point of Unnamed’s analysis was based on knowing that family size could confound things and so only comparing siblings within each family size. Tell me another method of analysis you’d prefer and we’ll do it and see what happens.

      • JayMan says:

        Let me start with a fundamental flaw in most social science, even the “genetically informed” studies:

        https://twitter.com/AmirSariaslan/status/531888996268650497

        You can’t “control” for this or that in studies, even those that use twin or adoptees, by statistically adjusting for this or that characteristic about the subjects. That’s the whole reason we cooked up behavioral genetic methods in the first place. The studies you mention, particularly that one you linked to, are chock full of useless nonsense like that.

        So yes, these guys:

        Studies done by Plomin, Turkheimer, etc. are not going to be making rookie mistakes.

        …do indeed make their fair share of errors, in part because they are working from a paradigm that’s flawed. If that sentence were true, I wouldn’t have much to blog about, especially in the post I linked to.

        Hence this:

        Having found that the studies which investigate genetic versus environmental effects find that it is mostly environmental

        …is nonsense.

        To briefly explain why, the problem is that (aside from the obvious genetic confounds in that standard social science study you linked to) even in genetically informed studies, you can’t assume an event in a person’s life is causal to some other later event: the characteristics of the person in question could cause both. This is true even if you “control for genes” by using identical twins, since “identical” twins are hardly identically, either phenotypically or genetically.

        If the adoption studies had found that divorce in childhood (of adoptive parents) was linked to adult problems, that might have helped (but would hardly be conclusive). But none have.

        However, what they found was that there was minimal difference between the size of the divorce effect in monozygotic twins versus dizygotic twins. Although one would not expect the non-genetic effect to be zero, because of the spouses, one would expect it to be more significantly less.

        Except was it significant? They didn’t give confidence intervals or even p-values, so they could just be reporting noise.

        Again, give Turkheimer some credit here for not running a completely useless study.

        You don’t know me very well, yet (understandably). 😀

        The entire point of Unnamed’s analysis was based on knowing that family size could confound things and so only comparing siblings within each family size.

        The commenter here mentions a key problem with that: age. The readers there – or at least the ones that complete surveys – are likely to be older. Older readers are likely to be skewed towards earlier borns.

        I’d argue that the effects of divorce are settled and don’t necessarily warrant further study: the effect is clearly non-existent, for if it wasn’t, it’d turn up in the shared environment of various behavioral genetic studies (which look at adults). It does not. This broadly rules out any aspect of the family environmental having much of an effect.

        (And before anyone brings up that different children in the same family experience divorce differently, think about it: the point of the “shared environment” measure is to capture aspects of the environment that children growing up together share – that’s the key word. This doesn’t mean such children will have identical environments, only that there are things they will share that children who didn’t grow up together will not. This includes divorce, which will have some commonalities to all children in the household – especially twins who are the same age. Yet no such effect is found.)

        • oneforward says:

          Older readers are likely to be skewed towards earlier borns.

          Are you suggesting that the general population would show similar numbers? I don’t see how the effect could possibly be that strong. For population growth alone to skew the numbers so that 75% of adults with one sibling are the eldest requires something like 6% growth per year and siblings generally separated by ~10 years. With more realistic growth of 1.5% and siblings separated by ~3 years only ~52% of adults with one sibling should be the eldest.

          Younger siblings could also be more likely to die as children, but this also can’t be a strong enough effect. If 52% of the population including the dead are eldest siblings, at least 30% of the population must be dead youngest siblings for 75% of the living to be eldest siblings. That is not the case for ages 22-31.

          Is there another reason the distribution should be massively skewed this way?

  18. Morgenstern says:

    I can think of one confounder no-one seems to have mentioned: whether the marriage was high conflict or low conflict before the divorce.

    (Edit: This sounds similar to comparing how stressful the divorces are but is actually different. Speaking from experience, low conflict marriages can still end in highly stressful divorces although presumably the reverse would be uncommon.)

    According to the research on marriage and divorce I’ve read, passed through the filter of me being a non-expert who mainly reads the abstracts, it looks like there are two kinds of marriages which end in divorce. High conflict marriages, involving abuse cheating and generally the sort of thing you had to prove to get a divorce before no-fault divorce became an option, make up about 20% of divorces. Low conflict marriages, which are clinically indistinguishable from successful marriages until the lawyers come out, make up 80% of divorces.

    Without a clear breakdown it’s hard to tell which, if either, is driving the phenomenon. I can think of four ways this could work off the top of my head, in order of plausibility to me;
    1. Kids from the 20% HC are doomed anyway, but those from the 80% LC are harmed by the novel stress of squabbling in the divorce.
    2. HC kids benefit from their parents divorces but are counterbalanced by the harm to LC kids.
    3. Kids in either situation are harmed to the same small extent.
    4. Most of the harm comes from HC kids, with the sheer number of unaffected LC kids serving to mediate it.

    These imply fairly different solutions, so ithat would be interesting to see which if any were the case.

    • Troy says:

      Thanks; this is helpful. I suspect that your option 1. is about right. In fact, I wouldn’t be surprised if some in the HC situations are actually benefited by divorce. But the relative rarity of this is important to keep in mind.

    • JayMan says:

      Non-paternity rates are closer to 1%.

    • Christian says:

      Since you say you’re reading around the subject..

      Whether or not divorce causes lasting damage, it certainly causes current and often lasting pain to the person who gets dumped and to the children. (To follow Harris’s explanation of why we should go on loving our kids.) The overall cost in pain to them is surely often higher than the pain avoided by the dumper.

      The biggest predictor of who will file for divorce is who will get the children. (Why most divorce filers are women -http://www.unc.edu/courses/2006fall/econ/586/001/Readings/Brinig.pdf )

      In a divorce, if who got the children (and assets needed to keep them) was always first allocated purely by chance between the parents, and then followed by a negotiation if desired, it would raise the expected cost to the partner considering initiating the divorce and hence inhibit them.

      If you’re not prepared to lose your children rather than stay in the marriage, why should you be able to impose that fate on your partner?

      • Matthew says:

        Because your partner is abusing the children as well as you?

        • no one special says:

          [Citation fucking needed]

          (Trigger warning: I’m triggered.)

          What proportion of divorces do you think feature spousal and/or child abuse? Be sure to discount divorces where the abuser files.

          • call_me_aka says:

            I’m really confused by this comment. Certainly abuse happens, and Christian’s suggestion is not, prima facie, a good solution in those cases…

          • Matthew says:

            It’s taking quite a bit of effort here not to release a full profanity-tirade on you, given that you started in with the swearing first and I had to file for an emergency custody hearing before even getting to my divorce from my abusive ex-wife, who was also abusive to my children.

            I don’t know what you’re triggered about. Your idiotic response has certainly triggered me, though.

          • no one special says:

            Matthew: I’m sorry. I went off half-cocked based on misunderstanding what you were saying. I have similar history.

            I accidentally pattern-matched your post to the argument “women file for divorce more then men because their husbands are abusive.” I then proceeded to be an ass based on that incorrect assumption. mea culpa.

  19. Tarrou says:

    Anyone know if any of these studies actually DNA match the “genetic” children to the parents? This may be my extreme cynicism at work, but it seems to me that allowing such a basic human tradition as cheating to gum up scientific work like this would be pretty irresponsible.

    The non-paternity rate I see bandied about in other research is around ten percent, which would be massive. More recent research puts the conservative estimate between one and three percent, easily enough to wreck a relatively small genetics study.

    • Anonymous says:

      The non-paternity rate is 1-2%. A large unbiased sample of Germany shows that the current non-paternity rate is 1%. The Sykes lineage shows 1.3% rate over 700 years. Every study that does not specifically select for people with doubts shows very low rates. Perhaps as high as 3%, but usually 1-2%.

      • Tarrou says:

        So you reiterate my point. Between one and three percent. I’d think Germany would be the low bound, there’s no way the US is less than that. That is still enough to bias the results of a small sample-sized genetic study with effects this weak.

  20. The Terman Study found that divorce lowered longevity by 5 years on the average, and a parental death did not have that effect.

    Since some of the subjects did alright after a parental divorce, this means that the ones who were affected badly lost more than 5 years.

    The negative effect was strongest in boys whose parents seemed to be on good terms before the divorce.

    All of this is from The Longevity Project

    Finding that divorce is likely to be bad for children isn’t contrary to the nurture assumption– I thought the assumption was that parental behavior doesn’t have much effect in the absence of trauma.

    Scott, have you taken a look at the Adverse Childhood Experience Study? It’s the one which found that various traumas had surprisingly large effects on health and longevity.

  21. “…pairs of identical male twins who married pairs of identical female twins…” Wasn’t that the basis of the Patty Duke Show?

  22. gwern says:

    > I’m sorry, I don’t have a copy with me

    Libgen is your friend!

  23. Q says:

    “So it’s not completely ruled out by the data that the short-term effects of divorce are robust and environmental, but the long-term effects of divorce are spurious and/or genetic. But this seems kind of like fighting a rearguard action against the evidence.”

    No, this seems totaly in line with all other hereditarian stuff I have been reading. For instance, intelligence seems to be more environmental at first, and more genetic later.

  24. Typhon says:

    I’m surprised by the title ‘Dark Side of Divorce’. Is there a light side to divorce ? (From the children’s perspective, I mean).

    Typhon

    • Nornagest says:

      Manipulating your way into extra Christmas presents from parents with a sudden pressing need to prove they still love you?

    • There can be. Some marriages are sufficiently high-conflict as to make the children worse off than they would be if the marriage were ended.

      • Tarrou says:

        Some = ~50%?

      • houseboatonstyx says:

        Some marriages are sufficiently high-conflict as to make the children worse off than they would be if the marriage were ended.

        Yes. Or, depending on how the ‘high/low conflict’ is measured, a child may be better out of even a marriage scored as ‘low conflict’, depending on what the conflict was about or what situations the conflict caused.

      • Anthony says:

        What are the odds that the custodial parent won’t then subject the kids to more high-conflict relationships after the divorce? With partners who have no pre-existing bond to the children?

    • nydwracu says:

      Yes, the light side is that it is very important for people to have the freedom to get out of contracts for no other reason than that they feel like it, never mind that 1) they are contracts 2) this will cause significant distress to the children in all but the worst cases (and the worst cases ought to be considered breaches of the contract anyway), and that distress lasts for at least ten years and probably forever.

      I realize that it is possible for conventional wisdom to be wrong, and that it is therefore a good idea to investigate such claims as “divorce is harmful” and “two-parent families are important”, but my parents divorced, one pair of my grandparents divorced, and my grandfather disappeared over 40 years ago, and the answers to those two questions are damn straight and very.

      People just don’t get over that sort of thing. Not after ten years, not after fifty. I can’t even blame my father or my grandfather (especially not my grandfather; my grandmother was a thoroughly terrible person and he probably would’ve drunk himself to death otherwise), but the effects are there.

      • Mark says:

        Yes, the light side is that it is very important for people to have the freedom to get out of contracts for no other reason than that they feel like it, never mind that 1) they are contracts

        There are plenty of contracts, e.g., employment contracts, that are deliberately designed for either party to terminate easily and at will. It seems like you’re stepping into Worst Argument in the World territory by anchoring marriage to the very austere-sounding word “contract,” which just happens to calls to mind fairly harsh paradigmatic examples. Of course, this isn’t to suggest your utilitarian arguments against easy divorces are wrong, just your semantic one.

      • Whateverfor says:

        To be fair, Efficient Breach over Specific Performance isn’t some wacky idea that only applies to marriage contracts. Once you accept marriage as a contract, some form of divorce is inevitable.

        Everyone thinks “divorce is harmful” and “two parent families are important”, it’s whether those effects are so bad that it’s worth forcing two people who hate each other to be miserable together for the rest of their lives. Not that I think Marriage-2014 is a perfect institution, but there’s no free lunch here.

        • AR+ says:

          I’m not even sure what I think about marriage, but I’m pretty sure that a major motivator behind stronger marriage is that people who are actually stuck together would make more good-faith effort at getting along. As usual, the effect would be greatest at the margin, and not so much result in people who hate each other being forced together as a much large number of people who can basically get along for the most part continuing to do so, rather than split because there’s no longer any passion.

          I have trouble believing that a significant fraction of pre-no-fault marriages were characterized by mutual seething hatred. And this, also: to what extent does divorce-related hostility arise from the fact that the kids and marital assets are now something to fight over? I don’t know, but given how close you can be w/ your siblings right up until it’s time to split the inheritance, I suspect it’s a factor.

        • Tarrou says:

          There is a free lunch. If you’re female anyway. If you’re a male, marriage is going to cost you, and for the rest of your life, if you’re lucky.

          Perhaps if the entrance barriers were a bit higher to discourage unprepared adolescents from hopping in and the exit penalties were evenly distributed, we could eliminate the greater part of these problems.

        • nydwracu says:

          To be fair, Efficient Breach over Specific Performance isn’t some wacky idea that only applies to marriage contracts. Once you accept marriage as a contract, some form of divorce is inevitable.

          Right. But there’s a difference between ending the contract because one party failed to uphold it and “either partner can, at any time, for any reason or no reason at all, break off the marriage, call in the lawyers, and loot the other.”

          There are marriages that are worse than divorce, but I have a hard time believing that 40-50% of them are.

          (There are more divorces than functioning marriages in my family. And divorce vs. functioning marriage correlates perfectly with atheist vs. Christian: all the atheists are divorced, and none of the Christians are.)

          • vV_Vv says:

            Right. But there’s a difference between ending the contract because one party failed to uphold it and “either partner can, at any time, for any reason or no reason at all, break off the marriage, call in the lawyers, and loot the other.”

            Pretty much all forms of legally valid contracts that regulate relationships between individuals or firms allow some form of unilateral termination.
            Granted, the terminating party doesn’t get to “loot” the other one, and in fact they typically have to pay some penalty.

            I agree that no-fault divorce in Western countries could use some revision, not a divorce fee but the terminating party should at least receive an unfavourable treatment in the distribution of assets.

            I don’t agree to a ban of no-fault divorce, however. Forcing somebody to have a relationship with a person they don’t like looks like a great cost. Moreover, if fault divorce is allowed, it would invite false accusations and other fraudulent behavior.

          • Whateverfor says:

            I don’t really oppose stronger marriage than Marriage-2014, I just don’t think the contract metaphor really supports it. When you start thinking in terms of contracts you’re going to end up with something like no-fault divorce, you want a different model. As the old model isn’t really available to us anymore, reducing divorce rates is not an easy problem.

          • Jaskologist says:

            Marriages aren’t arranged anymore; nobody is forced to enter a relationship with someone they don’t want. Choose wisely.

          • Lizardbreath says:

            “There are more divorces than functioning marriages in my family. And divorce vs. functioning marriage correlates perfectly with atheist vs. Christian: all the atheists are divorced, and none of the Christians are.”

            Moving from your first to your second sentence, you went from subtly referencing, to eliding, *non*functioning marriage. This suggests (not “begs”; that’s *improper use dammit*; I’m old 😉 ) the question: do all those married Christians have *functioning* marriages?

            There are very few divorces in my family. The two divorced couples in my generation were and are serious Christians. (In one, the woman left claiming abuse; in the other, the man left for someone new.) The couple in my parents’ generation were churchgoers; they stopped speaking to my parents before I was born (and before the divorce), so I don’t know how serious they were. The one in my grandparents’ generation were serious Christians (baby boomers often say the GI generation were mostly “just churchgoers”; not this couple).

            My parents are an atheist and an agnostic. Their marriage is unhealthy in a way that harmed kid!me*. I’m pretty sure anyone looking from the outside would consider it “functioning.” They only ever fought about what to do with me.

            So IOW, in my family, all the divorces were serious Christians. :shrug:

            *(My mom’s family is full of grade-skippers, my dad is the only one in his; my mom “recognized” me and knew what I needed, but didn’t have the “umph” to stand up to my dad; meanwhile he acted like a normal person and tried to “make me normal.” As a kid I used to beg my mom to divorce him.

            Around age 20 I decided that if *either* of them had had a free hand, it would’ve wound up with my developmental needs being recognized and addressed: if it had been my mom, great; if it’d been my dad, he’d have pushed me into a breakdown, after which, I think, he’d have realized my differences and needs were real. The problem, I decided back then, was the constant arguing with neither one ever getting to actually try out their plan. I haven’t revisited that opinion, don’t know if I’d still decide that now–but I was amused to read my own words and think “hey, like politics today!” 😉 )

        • Deiseach says:

          But is divorce all about “two people who hate each other and will be miserable for the rest of their lives”? Wasn’t that the whole point of ‘no-fault’ divorce – that there wasn’t sufficient poor or blameworthy behaviour by one party to get a divorce, so it was changed to make the contract terminable at will? And certainly there are cases where A is asked for a divorce by B, A thought the marriage was okay and wants to try counselling, mediation, etc. to make it work, but B just (for whatever reason) wants out?

          There are certainly marriages and cohabiting partnerships where the best thing for the kids and the other partner are to cut the other loose or get out themselves. There are also marriages and cohabiting partnerships where one partner, for whatever reason, decides they don’t want to be around any more, and it’s not because their partner drinks/takes drugs to excess/spends all the money/beats them and the kids/is cheating on them.

          • Matthew says:

            Wasn’t that the whole point of ‘no-fault’ divorce – that there wasn’t sufficient poor or blameworthy behaviour by one party to get a divorce, so it was changed to make the contract terminable at will?

            That may be the point, but just because a divorce was legally no-fault doesn’t actually mean a marriage didn’t involve abuse, adultery, or abandonment. Emotional abuse is really difficult to prove in general. Adultery, at least in Virginia where I live, requires hiring a private investigator and obtaining their documentation to meet the legal burden proof (imagine being in a marriage where the other spouse controls the finances and trying to arrange that).

            (I was both severely emotionally abused, and unfortunately unaware of the fact that neither pages from a diary explicitly discussing sex with other men nor copies of sexually explicit text messages could get past Virginia’s definition of hearsay, so I had a no-fault divorce despite both abuse and adultery.)

      • Lizardbreath says:

        My grandfather disappeared ~65 years ago. Then, ~25 years ago, when the hospital thought he was dying, they tracked down my father in order to ask him what to do. Thus I met my grandfather.

        It emerged that my grandmother’s family had blackballed him, though for the very good reason that at the time he was physically abusive toward both her and the kids. My father was nevertheless harmed by growing up believing his father didn’t care about remaining in contact.

        In any event, my father and his sister were…confused, perturbed, weirded out, whatever…at how, in the intervening time, he had gone from the monster of their memories to a harmless old man. As near as I can make out, their memories weren’t wrong; he simply…well, got old.

        (Me, I was slightly hurt that he had remarried, adopted her kids who weren’t biologically his, and now *their* kids were his grandkids, I just saw him every so often. Kid!me was all, “*I* am his rightful grandkid, not them!” I never said this to anyone, and I never took it out on those other kids; but I thought it.)

    • Scott Alexander says:

      It’s a pun. (“The dark side of the force”). I figured people would get it, but my focus group of Ozy didn’t, so maybe it just went over everyone’s head.

      • *slaps forehead*

        That’s not even a difficult or obscure pun. I really, really should have gotten that, and yet somehow I completely missed it.

      • suntzuanime says:

        Oh I just assumed you were being ironic about how you thought maybe divorce wasn’t so bad, even though conventional wisdom was that it was bad, but now you’ve discovered that it is actually bad, maybe, so you were sort of lampshading the whole “I’m about to tell you something everyone already knows” thing.

        EDIT: I don’t mean this as a slam on your article, my impression is that your readership tends toward the clever contrarian, and they can often use a little reinforcement for the correct (maybe) bits of conventional wisdom.

      • Deiseach says:

        I think one reason I don’t much like puns is because I’m too literal-minded to read something as a pun. It really is not funny when the other person has to painstakingly explain to me “No, see, that was meant to be funny because…”

        I’m a bit like Drax from “Guardians of the Galaxy” in that respect 🙂

        Rocket (the raccoon): These people are completely literal. Metaphors are gonna go right over their heads.

        Drax: Nothing goes over my head. My reflexes are too good.

      • Typhon says:

        Well it’s a consolation to see that even native English speakers didn’t get it.

  25. Anonymous says:

    As far as I remember, The Nature Assumption was mostly about refuting studies that showed shared environments were important. The separated twin studies were really the only evidence the author presented that showed the importance of genetic effects.

    I came away from the book feeling like we mistakenly rejected the null hypothesis (which is that parenting doesn’t have an effect on a child’s outcomes). Future studies could show that social effects were more powerful, but the ones we had already done didn’t show that since they didn’t control for genetics.

  26. Matthew says:

    Thank you for this post. It’s made me feel better than anything I’ve read in quite a while. Which might seem odd, given that I’m divorced and have children. But I’m far more terrified of The Nurture Assumption being entirely correct than I am of any lingering effects on my children from divorce (which are unquestionably less than the effects of staying in an abusive environment would have been, anyway).

  27. lambdaphage says:

    > “children of divorce scored lower than children in intact families across a variety of outcomes, with the median effect size being 0.14 of a standard deviation,” this last clause of which is almost New Cuyaman in its agglomerativeness.

    Seems reasonable to me. In fact, it just sounds like the author was trying to paraphrase the definition of Cohen’s d. It’s natural to scale effect sizes by the standard deviation, which is the expected deviation from the mean. Effects which are much smaller than the standard deviation are often neglectable, since they are dominated by noise. It says, in effect, “yes, children of divorced parents tend to do worse on average, but only by about a seventh of the amount that children’s outcomes tend to vary just by chance.”

    • Cyan says:

      Indeed. Perhaps it will be more understandable expressed in IQ-point-equivalents: if “the median effect size being 0.14 of a standard deviation” were describing an effect on IQ, one could equally say that the median effect size was 2 IQ points.

  28. Patri Friedman says:

    But does this mean that the claims about other, similar effects, in The Nurture Assumption are wrong? I see a steady stream of news stories about sociology papers finding correlations between parent & child, and every now and then I go find the paper, and they almost never even mention the possibility of a genetic causal link, let alone correct for it. Combined with the strong data from twin studies on lots of things that aren’t divorce, the book’s basic thesis that there is reams of completely erroneous research that ignores heredity still seems true in general, even if it isn’t alway true in specific.

  29. G Joubert says:

    When it comes to birth order, it’s most evident when it comes to the eldest female child, especially when she is also the oldest child overall. You can’t miss her. Mama’s little helper in every way imaginable. And then she grows up to be that as an adult too.

    As for the different ways daddy is perceived in homes headed by a single mother, widows versus divorcees, it’s really very easy to see and grok. In intact “happy” families where daddy is killed, he becomes a sacrificed hero, whereas in instances of a broken home, daddy gets slimed.

  30. Anonymous Coward says:

    Jay Joseph has a book coming out next month critiquing many of the assumptions behind twin studies:

    http://www.routledgementalhealth.com/books/details/9781138813069/

    In his previous books, he focused on the studies of twins reared together. In this book, he also tackles studies of twins reared apart, including the famous Minnesota study.

    In addition, James Flynn has told me via email that he is also finishing up a new book which puts a new spin on twin studies, but couldn’t divulge more info than that. Whatever the case, it looks like the debate is far from over.

  31. Steve Sailer says:

    The college admission process has become so competitive and complicated that I suspect children in two parent homes have a growing advantage.

  32. Pingback: Genocea Up on Positive GEN-003 Data for Genital Herpes – Analyst Blog | Herpes Survival Kit

  33. Pingback: Lightning Round – 2014/11/19 | Free Northerner