Stop Confounding Yourself! Stop Confounding Yourself!

As a perk of my job, I get a free subscription to the American Journal of Psychiatry. I am still not used to this. No enraging struggles with paywalls. No “one year embargo on full text”. I just come home and find all of the latest and most interesting journal articles have been shipped directly to my house. Modern technology is truly amazing.

Its latest is Takizawa et al’s Adult Health Outcomes of Childhood Bullying Victimization: Evidence From A Five-Decade Longitudinal British Birth Cohort. It has since been picked up by Fox, the Washington Post, and even Xinhua. I think that’s enough to qualify for “made world headlines”.

The study took some British kids in 1958, sorted them by how much they got bullied, and checked how they did forty years later. In fact, the frequently bullied kids had nearly twice as much psychiatric disease, were twice as likely to attempt suicide, were twice as likely to drop out of high school, and even had double the unemployment rate. Worse physical health, worse cognitive function, less likely to get married, et cetera, et cetera.

Those must be some bullies.

But correlation is not causation. There’s an alternative possibility. Maybe bullies only pick on unpopular disadvantaged kids. And maybe these kinds of things are stable, so that unpopular disadvantaged kids are more likely to grow up to be unpopular disadvantaged adults. The sort of adults who are more likely to have psychiatric disease, drop out of school, be unemployed, et cetera. That sure sounds plausible.

So the researchers “controlled for confounders”. They used a scale called the Bristol Social Adjustment Guide to figure out how socially well-adjusted the kids were, then added in their social class, their family’s level of contact with child protective services, their IQ, their attractiveness, and even how much their parents loved them (really! check the study!)

They controlled for all these things and found that the bullying-outcomes link was still robust. They concluded that this meant their finding wasn’t just that bullies were bullying kids with problems, it was that bullies were causing the damage themselves.

Do you believe that? It all comes down to one question.

Who is better able to look deep inside you and judge the mettle of your soul? A playground bully? Or the Bristol Social Adjustment Guide?

My money is on the bully. Bullies are like sharks: horrible pinnacles of evolution. Animals have been learning to navigate social dominance hierarchies through violence since pecking orders in chickens, on through wolf packs and chimpanzees, and up into humans – and we are very good at it. The bully is the purest manifestation of the primal instinct, which is why he crops up untaught and unbidden in near-identical form in schoolyards from Los Angeles to London to Lanzhou. And like sharks, a good bully should be able to smell blood in the water and know when an opportunity to attack presents itself.

Most of the findings of this study were in the “frequently bullied” population, and part of the criteria for “frequently” was bullying both at age 7 and age 11. Unless that’s just one really persistent guy, that means the child has gotten independently selected for targeting in two different environments. That could be bad luck but could also be the effect of high inter-bully reliability in what (persistent) qualities make a good victim.

So let’s take another look at those confounders we supposedly controlled for. Where’s height? You think short kids are bullied more often than tall kids? I do. Height is closely related to career success, to attractiveness to the opposite sex, increased happiness and self-esteem, and decreased psychological morbidity. This is something every bully knows intuitively, but which the Takizawa study didn’t think of and therefore couldn’t control for.

But it’s giving them too much credit to be bringing in weird stuff like height-mental-health correlations. What about social skills? Yeah, sure, they did that Bristol Social Adjustment Guide. I’m looking at it right now, and it’s asking the students’ teachers to rate items like “hostility towards adults” and “depression”. I don’t believe that teachers filling numbers into hokey little boxes can capture an assessment of a kid’s social skills as well as a bully trying to decide who can safely be picked on can.

So I will come out and say it: I do not trust the practice of “adjusting for confounders”, at least not the way this study does it. You are adjusting for an imperfect measurement of the confounders you can think of. If you find that there is lingering correlation, then either your hypothesis is true, or you didn’t adjust for confounders well enough. Given extraordinary results, like being bullied at age seven making you 25% less likely to be married at age fifty, the “you didn’t adjust for confounders well enough” option starts to look really good.

I think the proper way to do this study would have been to do an anti-bullying intervention at a couple schools, leave a couple similar schools as controls, and if the anti-bullying intervention successfully decreases bullying, compare outcomes for children at the two schools. I understand this probably would be logistically impossible, plus you’d have to wait another forty years. But given that you cannot do the study right, I am not sure that doing the study this way adds anything, except of course widely-read articles in every news source in the world.

I would also compare to Reming et al, which attempts much the same study and finds no association after adjusting for their confounders of choice (which, oddly, are much fewer than in the current study). They also find that parent reports about bullying (the method Takizawa et al used) are wildly unreliable, with an inter-rater agreement of just 0.11 with reports by teachers or the children themselves (the statistic goes from perfect agreement being 1.0 to zero information being 0.0). For a completely false measure of bullying to find such spectacular effects is really suspicious, and now we need to consider not only the differences between the types of kids who are and aren’t bullied, but the differences between the types of parents who do and don’t think their kids are being bullied.

Since I insisted on giving this post a silly title, I will now share with you the most interesting perspective on psychology and the “stop hitting yourself” phenomenon I have read all week. This is from Jonathan Haidt on Kohlberg’s moral stages:

During elementary school, most children move on to the two conventional stages, becoming adept at understanding and even manipulating rules and social conventions. This is the age of petty legalism that most of us who grew up with siblings remember well (“I’m not hitting you. I’m using your hand to hit you. Stop hitting yourself!”). Kids at this stage rarely question the legitimacy of authority, but learn to maneuver within and around the constraints that adults impose on them.

I always just thought that was a really dickish joke. I didn’t realize it had a deep philosophical underpinning.

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

92 Responses to Stop Confounding Yourself! Stop Confounding Yourself!

  1. suntzuanime says:

    This doesn’t even touch on the incredible mischief that you can get up to if you deliberately pick which confounders you control for in order to make the results come out how you want. I’m not suggesting that the study in question did this, but no discussion of how controlling for confounders is total bullshit is really complete without a mention.

    • suntzuanime says:

      This can even happen not-legitimately-but-close-enough-that-a-hypocrite-human-can-sleep-at-night via motivated continuation. If your study doesn’t turn out how you want (i.e. how is OBVIOUSLY TRUE), that’s clearly a sign you haven’t controlled for enough confounders, so you slap another confounder on the pile and run the numbers again. You keep doing this until you finally get the obviously true result and you know you’ve finally controlled for all the confounders and can stop.

  2. JTHM says:

    In my experience, school administrators’ anti-bullying efforts are completely counter-productive. If your classmates constantly berate you and beat you up in the hallways, and you tell a teacher, that teacher can either simply take your word for it and punish the aggressors, or investigate to see what truth there was to the claim and then punish accordingly. If they were to simply take you at your word, a war of Accuse Your Enemies Before They Accuse You would erupt, and everybody would simply use the administration itself as a way to bully. And to no small extent, this really is done. Teachers could potentially face consequences for inaction if faced with pressure to clamp down on bullying, and so prefer to deal out uninformed punishment rather than stay neutral. But what of those who investigate? They rarely have anything more than student testimony to rely upon, and bullies call upon their friends—of whom they have seemingly infinite supply—to perjure themselves in the bully’s defense. And so the administration declares that YOU must be the bully. Now in addition to your problem with schoolyard thugs, you have the teachers torturing you and telling you what a horrible bully you are.

    This happened to me in elementary school, and I eventually had to change schools to get away from the administration after it happened. But my problems with bullying from other students persisted until middle school, when I finally realized that violence was the only answer, and retaliated by injuring school bullies in front of large numbers of witnesses. Sure, I got suspended, but it was all worth it in the end, and I never had to fight anyone familiar with my reputation again. After making examples of just a few people, I had bullies lining up to make their peace with me. (In case that you worry that I became a bully myself, know that I never threw the first punch.) And what did the administration contribute to my efforts? Less than nothing. It was only the fear of retaliation from the system that prevented me from using violence earlier. Bullies can always get away with violence because of their control over witness testimony, so clamping down on violence merely disarms the innocent.

    So how is this relevant? If you want to see how a decrease in bullying affects the experimental group, you need to STOP the anti-bullying campaigns, not step them up. When students fight in the hallways, just break up the fight and let the combatants go about their business rather than punishing them. If the bullies know their victims can strike back without facing the wrath of the administration, they will exercise at least some measure of self-restraint.

    • Ran GTPase says:

      I don’t know which of awful counterproductive administrative interventions and awful useless FOR THE LOVE OF GOD WE HAVE A BETTER UNDERSTANDING OF CAUSALITY THAN YOU ARE USING PLEASE JUST STOP studies make me more angry. And in this particular case there’s a positive feedback loop between them!

      • Vanzetti says:

        >>>How can someone live all the way to fscking adulthood and NOT UNDERSTAND THIS?

        How many children who were bullied go on to become school teachers (as opposing to shuddering at a mere thought of having anything to do with a school)?

        • Ran GTPase says:

          Sorry, I edited out that section of my post (since I was unhappy with how rhetorical the question was) before my browser loaded your reply. But I don’t think people who weren’t bullied (for example, me) have a very good excuse for not figuring out that you usually can’t significantly change power/status differences with simple rules (since the previously powerful can usually just game the rules more effectively than the previously weak).

        • Sniffnoy says:

          You’re making a number of non-obvious assumptions here. Which may be true, but you shouldn’t expect them to be obvious to other people who are unfamiliar with the subject. E.g.:

          * Bullying is about status and social power rather than brute physical force
          * Being able to hurt under one set of rules correlates with being able to hurt people under a different set of rules

          I mean, that last one you just assert: “The previously powerful can usually just game the rules more effectively than the previously weak.” That may be the case, but why would I expect it to be so? Is there a “positive manifold of power” or something? Someone who thinks of bullies as dumb brutes is certainly not going to expect this.

        • Ran GTPase says:

          > Being able to hurt under one set of rules correlates with being able to hurt people under a different set of rules

          I would expect the correlation to get weaker the more well-designed, well-enforced, and different-from-the-old-rules the new rules are. But I was careful to say “simple rules” before, and the thing I expect everyone to be able to recognize has less to do with the details of bullying than the common uselessness of rules which *aren’t* carefully designed and enforced.

        • peterdjones says:

          Teachers can be bullied too.

        • Sniffnoy says:

          > the common uselessness of rules which *aren’t* carefully designed and enforced.

          Well, I’ll certainly agree that everyone should recognize that!

        • Ran GTPase says:

          > Well, I’ll certainly agree that everyone should recognize that!

          And that tautologousness’s a big part of why I removed it from the post in the first place!

        • Most bullying doesn’t seem to involve physical force?

    • ozymandias says:

      I also got bullied and sexually harassed a lot, but I am incredibly wimpy (also, most of my bullying was non-violent– I probably could have gotten away with hitting the people who groped me but not any of the others). I found the only effective way of preventing them was to befriend people who were demonstrably willing to threaten and punch people in my defense. Although that failed when I ended up in a class where I didn’t know anyone… :/

      • nydwracu says:

        (I tried to leave a longer comment, but it got eaten.)

        Non-violent bullying is a lot harder to escape than violent bullying, especially if it’s frowned upon to escalate it to violence, and especially if there’s an administration or an inescapable power placed higher on the hierarchy that the bully can appeal to—bullies need to be skilled at intuitively navigating the social environment, and that happens to be the exact skill necessary for getting the administration to take one’s own side.

        (From this it follows that there are substantial negative effects to bans on dueling.)

        • Randy M says:

          Your comment about escalation reminded me of my daughters last night; eldest, 6, came in to say the middle, 4, tried to bite her. Seems that she was threatening to bite her to get her space, and in a way I was relieved to hear that it could escalate to only the “threat of force” level.

    • Thasvaddef says:

      Your “stop the all anti-bullying campaigns and let the kids get on with it” strategy would seem to merely shift the balance of power from the popular kids (who rule in the “punish based on witness testimony” strategy) to the strong kids, which isn’t the same as stopping bullying.

      • suntzuanime says:

        There will always be a dominance hierarchy. Bullying is a struggle to come out on top of the dominance hierarchy. It is especially common in children because they have no power over adults and cannot bully them into not recognizing their bullying as bullying.

        Anti-bullying campaigns are doomed to failure, a lack of anti-bullying campaigns is doomed to failure, everything is doomed to be devoured by the beast known as “society”. The best thing we can do to cut down on bullying is to make it easy to opt out of any particular social network. (Schools and prisons but I repeat myself are especially bad about this, which is another reason you see so much bullying there.)

        • Daniel Speyer says:

          Hierarchies can be based on something other than bullying. Communities that care about something beyond themselves tend to be low-bullying environments because the best path to status is being good at that thing. (See Paul Graham)

          This suggests that an effective anti bullying measure would be to provide things to care about. Granted, this would be hard.

        • Anthony says:

          Reading Daniel Speyer’s comment below, I think it’s more complicated. There are *multiple* hierarchies in a large social environment like a school. Bullying, by and large, is an attempt by those who are not at the top of the academic hierarchy to make the bully’s hierarchy more “important”. Bullying isn’t always by the “dumbest” kids, but it’s rarely by the smartest, even when those kids otherwise have the advantages which would let them be successful bullies.

        • Anonymous says:

          One good way to come out on top of the dominance hierarchy is to find a hierarchy you’re on top of and assert it over all the other possible hierarchies.

          In this sense, the studious (I don’t like using “smart” here) kids who want the focus of school to be on homework, which they just so happen to be good at, are at least as much bullies as the non-studious kids who are trying to avoid being marginalized by pushing the focus to social interactions instead. The non-studious kids have a better claim to justice, since homework is unnatural and immoral and imposed by tyrants from outside. Studious kids are collaborators who are getting what they deserve.

        • AJD says:

          In this sense, the studious (I don’t like using “smart” here) kids who want the focus of school to be on homework, which they just so happen to be good at, are at least as much bullies as the non-studious kids

          “Gentlemen, you’ve both worked very hard, and in a way, you’re both winners. And in another, more accurate, way, Barney is the winner.”

        • Roxolan says:

          “the studious kids who want the focus of school to be on homework, which they just so happen to be good at, are at least as much bullies as the non-studious kids who are trying to avoid being marginalized by pushing the focus to social interactions instead.”

          Whether they both fit into the definition of “bullies” is an argument not worth having. They’re not on equal moral ground though. Studious kids mostly don’t cause misery on purpose, nor in the same amount.

        • Anonymous says:

          Morality gets so much simpler when you only count misery you inflict on purpose! All you have to do is blind yourself to the consequences of your actions, keep your head down, and do your homework.

          And if anyone tries to get you to stop causing them misery, *they’re* the ones in the wrong, because they’re doing it on purpose, and purpose is the root of all evil.

          I guess this is what the SJW types mean when they talk about unexamined privilege.

        • Alexander Stanislaw says:

          Anonymous, are you trying to explain the mindset of a bully, or are you arguing that said viewpoint is correct?

        • Anonymous says:

          Children do not have the philosophical sophistication to explicitly understand the nature of the social dynamics at play here; they simply feel themselves being wronged and intuitively respond. I am trying to explain the situation as a whole. The situation as a whole includes the mindset of the bullies, but people’s explicit beliefs are less important for social dynamics than most people seem to think.

        • ozymandias says:

          Anonymous, could you please give a specific example of how smart children cause as much misery as bullies? If it is true that homework is an unjust system*, I am still unclear about why we are blaming studious children for it, because even if every studious child in the world stopped doing their homework homework would still be assigned. I also think the smart/studious distinction is important: I got shit grades and never did my homework, but I was perceived as smart and bullied for being smart.

          *I am unclear if your position is pro-unschooling, which I am heartily in favor of, or anti-learning, which I am not.

        • One of the variable is how badly people at the bottom of the hierarchy are treated.

        • ozymandias, I think there’s a description in Born to Kvetch of students who were good at Talmud bullying those who weren’t, but I don’t have the book handy.

        • Oligopsony says:

          If it is true that homework is an unjust system*, I am still unclear about why we are blaming studious children for it, because even if every studious child in the world stopped doing their homework homework would still be assigned.

          If academic success is a positional good, then everybody benefits from not spending real resources on it.

        • ozymandias says:

          Nancy: That sounds plausible but also meaningfully different from what Anonymous is talking about. (Maybe? I’m not sure what their claim is.)

          Oligospony: I am working from the premise that some people being studious is a good thing– either because it actually teaches them valuable skills or because it signals hardworkingness. I can certainly see an argument that punishing kids who set the curve is good for the other students, even if it is not good for Society As A Whole, but that still leaves Anonymous’s equation of scholarship and bullying somewhat puzzling to me.

        • I’ll throw in a stronger hypothesis– that bullying is to a large extent, based on what the adults care about.

      • JTHM says:

        Ah, but you see… It’s difficult to know who can effectively retaliate and who cannot. If you’re fortunate enough not to have been in many fights, you’d be surprised how hard it is to predict how dangerous someone is to provoke based merely on physical appearance. I myself was considerably less physically fit than all of my opponents, and they all judged me safe to assault. Bully enough people who *seem* like wimps, and the odds that one of them could take you in a fight become overwhelming very quickly.

      • nydwracu says:

        Are the bullies the strong kids, the ones who could pick on anyone and have a good chance of winning—or are they only strong enough to have enough of a supply of kids they know they can win against?

        I’ve heard plenty of anecdotal evidence for the latter “calculus of savages” theory, and none, as far as I can remember, against it. I also have the evidence of my father, who went on a business trip to NYC in the pre-Giuliani era, got lost, and wound up in the middle of Harlem. But he’s six and a half feet tall, so—and I mean this in the most causal way possible—no one fucks with him. I’m an inch shorter than he is, and I can cite the same thing: I don’t live in a good area, but I have no problem at all walking around at night. If I got into a fight, I would probably lose—but I’m tall enough that I appear to potential assailants as too risky to be a target.

        The calculus of savages is the minimization of risk: bullies don’t attack people who fight back. (Note the implications here outside the realm of the schoolyard: how much of progressivism operates on the logic of the grade-school bully? Valleywag is a perfect example of the calculus of savages.)

        • Anthony says:

          My middle-school experience somewhat validates this. I was subject to harassment and teasing, which eventually ended up with the “bully” starting a fight with me. I fought back. Not very effectively – to be honest, I “lost” the fight. But just the fact that I tried was enough to keep me from getting harassed by that kid for the rest of middle school. Interestingly, the crowd that formed around us, while mostly kids who were friendlier with the bully than with me, cheered for me, and some of them congratulated me on “winning the fight” afterwards.

    • chauvinistic celestial-undefined hetero-elitist bigot says:

      You’re right. For one, if you get rid of playground rules, bullying and mischief goes down: http://www.independent.co.uk/news/world/australasia/new-zealand-school-bans-playground-rules-and-sees-less-bullying-and-vandalism-9091186.html

      • CaptainBooshi says:

        I would just like to note that this article does not actually support the original poster’s proposed idea. The article talks about getting rid of the health and safety-based rules on a playground, originally to foster more ‘active play’ among the kids. As far as I can tell, the original poster was discussing the rules concerning interaction between kids, both in and out of school.

        The article itself is quite interesting, though, and I would like to see this replicated and expanded to see if it will continue to work. It also makes a lot of sense to me, that kids would be less bored without all the (increasingly stupid every day) restrictions on play, and be less likely to out act in destructive ways.

    • CaptainBooshi says:

      I have to strongly disagree with this statement: “If the bullies know their victims can strike back without facing the wrath of the administration, they will exercise at least some measure of self-restraint.”

      I think you are over-generalizing from your own experience here. It worked for you because you were both willing and able to fight back. Your strategy will just mean the bullying will shift to the kids who don’t have the will to fight, or kids weak or small enough that they can’t effectively fight back.

      I’m not supporting current anti-bullying campaigns, by the way. I agree that they can be quite terrible at their stated goals. I’m just saying your preferred solution would almost certainly be just as bad.

      • DavidS says:

        Agreed, although it’s not just a matter of the strongest: popularity is important because numbers are important, plus the removal of control from above doesn’t automatically lead to just fights to the death, so other things remain important: wit, ability to mock/defend, capacity for cruelty etc. etc.

        In terms of people discussing how you could run an experiment like this: the laissez-faire approach of kids sorting out their own affairs was by all accounts a popular approach in UK public schools (‘public’ here meaning the posher private schools). This tends to be presented either as giving people backbone, character etc. or as a source of immense, often systemised bullying and indeed physical and sexual abuse of younger/weaker students by older. Sometimes it seems to be regarded as both.

    • Anonymous says:

      Anecdote 1: I (male) was bullied in school. Also, one time, two girls started a fist fight in the hall and were dragged away by cops, so both genders were doing it.

      Anti-bullying campaign begins.

      10 years later, younger sister reports no (physical) bullying of anyone, and is shocked that physical violence ever took place in the school.

      Anecdote 2: The reason I never hit them back was not because I was afraid of consequences, but because I believed that it was wrong to strike someone unless they struck me first. So, of course, the bullies just casually hit me or kicked me once as they passed or tried to make it look accidental… making it difficult to pull off any justified and heroic retaliation. Only one time did anyone ever attempted an actual prolonged fight (to which I responded not by hurting him but by making him very uncomfortable, he left me alone after that 😉 )

      Generalizing from myself, other victims may be pacifist (or too physically weak to fight back), so your “just break it up” strategy would not have helped me.

    • Ialdabaoth says:

      In my experience, school administrators’ anti-bullying efforts are completely counter-productive. If your classmates constantly berate you and beat you up in the hallways, and you tell a teacher, that teacher can either simply take your word for it and punish the aggressors, or investigate to see what truth there was to the claim and then punish accordingly. If they were to simply take you at your word, a war of Accuse Your Enemies Before They Accuse You would erupt, and everybody would simply use the administration itself as a way to bully. And to no small extent, this really is done. Teachers could potentially face consequences for inaction if faced with pressure to clamp down on bullying, and so prefer to deal out uninformed punishment rather than stay neutral. But what of those who investigate? They rarely have anything more than student testimony to rely upon, and bullies call upon their friends—of whom they have seemingly infinite supply—to perjure themselves in the bully’s defense. And so the administration declares that YOU must be the bully. Now in addition to your problem with schoolyard thugs, you have the teachers torturing you and telling you what a horrible bully you are.

      In my more cynical moments, I’m reasonably convinced that this is all completely intentional. Adults have just as much motivation to bully kids – if not more – than other kids do. Most people can intuitively identify which kids are defective, and treat them accordingly.

    • Desertopa says:

      I’ve known a number of people who had similarly unfavorable experiences with school administrations dealing with bullying, but I can say from personal experience, and from the shared experiences of some other acquaintances, that it’s not always nearly that bad.

      It’s true that if schools follow the “investigate, then punish” route, then they usually don’t have any direct evidence of the bullying to work with. But *good* teachers and school administrators are not idiots. It doesn’t take a master of psychological profiling to keep up some level of observation of the relationships between students, and develop clear enough personality profiles of the students to have a good idea prior to any accusation who’s likely to be a bully and who’s likely to be a victim.

      This takes effort, and it seems that less hardworking school administrators, or ones whose time is too heavily taxed by other tasks, will often insist that the amount of work it would take renders it effectively impossible. But I’ve also known some very effective school administrators and teachers who regard it as a basic and necessary part of their jobs.

      • Deiseach says:

        What’s difficult is not physical violence – if someone starts a fight or hits another kid or has a habit of throwing chairs etc. around, you can see that and act on it.

        The tougher thing is verbal bullying, or even instigation. Anecdote time again: one kid where I worked was very easy to wind up and he’d throw chairs, throw punches, get in fights, the lot. He was the one who got in trouble for this.

        The one who didn’t get in trouble was the sly kid who egged him on because he wanted to disrupt the class or just for the amusement of it. Because although you know, I know, everyone knows that when Johnny (not real name) makes an ‘innocent’ remark about mothers or babies or fathers or homework (depending), this will set Tommy (not real name) off due to his home circumstances, when you challenge Johnny about this, you’ll get the “I didn’t say anything! I never mentioned Tommy’s mother! We were only talking about my aunt had a new baby!”

        Johnny knows Tommy’s home circumstances, we know Tommy’s home circumstances, he and we know why Tommy (who has behavioural problems) will go off like a rocket when you hit one of these sore points – but how can you prove that Johnny is the one causing trouble, when it’s Tommy kicking the door in?

        About the only consolation you have is that Johnny is too clever by half and will end up in jail in five years’ time when he gets caught for petty swindles and other con jobs, and that’s not really a consolation because who wants to know that Kid A is headed straight from school to jail?

        • Ialdabaoth says:

          About the only consolation you have is that Johnny is too clever by half and will end up in jail in five years’ time when he gets caught for petty swindles and other con jobs

          Not if he goes into business. I know a few con-artists who’ve run multiple successful investing scams, and who have spent 40+ years keeping one step ahead of the law. How’s that quote go? “The difference between investment crime and other crime is that the question is far less often ‘did he do it?’, than ‘yes, what he did was terrible and wrong, but was it illegal?'”

    • I have heard about anti-bullying campaigns which were structured very differently. The idea was to build an anti-bullying culture at the school rather than to punish bullies.

      Children were taught how to recognize bullying, that it was bad, and to be enforcers. (One consequence was that a teacher might be told to stop bullying.)

      I’ve heard that this approach works, but I haven’t heard of any studies. Anyone have information?

    • Deiseach says:

      Not bullied myself, but did work in a school where we had a lot of kids with behavioural problems.

      Anti-bullying (like anti-homophobia) interventions sound like a great idea – from the outside. From the inside you quickly realise this only gives the kids a whole new spectrum of tools to use (as in “Hey, I never knew before you could use that term/do this to someone!”).

      For instance, we got the information packs from a “Stop Bullying of LGBT” students from just such a programme, including posters we were supposed to put up with lists of slurs on them. The idea was “Teach children these are not acceptable”.

      We didn’t put them up because we knew the result would be “Bunch of 12-16 year old boys and girls learn new insults to pick on victims”.

      Re: one person’s word against the other, that’s why we have CCTV cameras in the corridors. Joe says Mike hit him, stole his books from his locker, or threw water all over him? Check the tape and see what happens.

      Also helps when Mike says Teacher X or Y shoved them around or slapped them, because we’ve had those accusations made and we’ve had parents steaming in literally shouting and screaming at the teacher about “You hit my kid!”

      If you can show that no, at 11.15 a.m. in the corridor outside the gym, Mr Brown did not slap little Timmy around the head, that saves a lot of trouble (though you still get “Are you calling my kid a liar?”)

    • Scott Alexander says:

      I see where you’re coming from with this, but I don’t think punishing bullies is as hard as you think. I would expect teachers to have a pretty good idea who the bully versus the victim is in a lot of interactions – even if they’re not very observant, stereotypes will get you a lot of the way. Is it the word of the tiny, shy, bookish student against the really popular huge jock? Most likely the former was the victim.

      There is something very stupid close to what you describe, which is punishing everyone in a fight including people who were just defending themselves. But I think getting teachers more involved in stopping bullying is likely to be a good thing, especially with halfway competent teachers.

      I would also support making everyone wear third eyes.

      • Ialdabaoth says:

        Is it the word of the tiny, shy, bookish student against the really popular huge jock? Most likely the former was the victim.

        Key word here: ‘popular’. It can be uncomfortable to punish someone who’s just so likable, especially for picking on an obnoxious little jerk that you wouldn’t mind harassing yourself if not for professional requirements.

        Also, typically that really popular huge jock will have a parent or uncle in some kind of social position where writing them up might have negative consequences, while the bookish little twerp will often come from a family with little to no social capital in the community.

        In the reverse situation (jock comes from the wrong side of the tracks, bookish kid is from a well-to-do family), the problem DOES get dealt with quickly – and in a way that fast-tracks the poor jock straight into the criminal underclass where he belongs.

        I would also support making everyone wear third eyes.

        http://thinkprogress.org/justice/2014/04/15/3427016/boy-with-disabilities-bullied/

        Somehow, I don’t think they’ll be very keen to implement that.

      • Considering that teachers seem to be extremely bad at identifying and punishing bullies, I’m going to assume there’s something you’re missing.

        • Scott Alexander says:

          Are you sure of this? The most memorable time I was bullied, I put up with it for a while, then told a teacher who took care of it pretty quickly. My experience might be anomalous, but it also seems possible to incorrectly go from “There’s still a lot of bullying around” to “telling teachers doesn’t help”, which seems as wrong as “There’s still a lot of disease around” to “going to the doctor doesn’t help”

        • suntzuanime says:

          Yeah, I too have had the experience of the school administration successfully dealing with my bullying. On the other hand, I was reasonably good at playing the right cards with the administrators, and I had parents that would have been willing and able to raise a stink if it came to it. So probably what they’re actually bad at is helping the helpless.

          Of course this was physical bullying. I doubt that any school administrator is going to be able to handle socioemotional bullying.

        • At this point, I’m sure of a different thing. Bullying seems to amplify the typical life fallacy.

          I see people remarkably sure that their experience is typical, whether it’s that there’s no hope if you’re bullied, or that drastic physical retaliation is the solution.

          You seem to have a somewhat milder version, but I’m willing to bet that you had more cred with your teacher than many bullied kids do.

      • JTHM says:

        It wouldn’t be hard for a moderately competent and benevolent school administration. Perhaps I just lived in a weird little bubble where no such thing was to be found, and the rest of the world enjoys school principals who have working brain stems and hearts that aren’t three sizes too small. But I come from a pretty posh area where the public schools ranked among the best in the nation, and the school employees I dealt with combined the intellect of Forrest Gump with the altruism of Gordon Gekko. If I had to bet, I’d wager that most other people had it worse.

      • nydwracu says:

        Then the bully learns to make himself look bookish. Overdogs taking on the appearance of underdogs as a bullying strategy is very well-established—especially after the bully leaves the schoolyard. That’s one thing Moldbug is right about.

      • b says:

        You don’t think there would be similar issues as with false rape accusations?

      • Sam Rosen says:

        I think guidance counselors need to be more involved in understanding the status dynamics of the students they work with. That way they can more accurately judge who is likely to be bullying whom.

        I don’t even think it would be that hard. You could set up a “Family Feud” type set-up where students are asked who people are dating, who people are friends with, and who is popular. The more your answers align with other people’s answers, the more points you get.

    • Doug S. says:

      I’ve read that most bullying is done by people in the middle of the dominance hierarchy. The people at the top are secure in their position and don’t need to, but the people lower down are at risk of falling further and try to make sure that the other people looking for targets have someone else to victimize.

      • Anonymous says:

        I agree that boys who bully are in the middle, but I’m not sure your explanation is correct. The situation with girls is very different. The most popular girls control gossip and use it to reinforce the pecking order, though this is rarely called bullying. Also, they might orchestrate the bullies in the middle against the victim of the day.

    • Mary says:

      You left out “punish you for not trying hard enough to get along.”

      You don’t even have to complain to get that one.

    • Walter says:

      Had a similar experience, minus the superheroics. Pre-anti-bully-program bullies called victims nasty names. Post anti-bully-program bullies called victims the same names, plus they called them bullies.

      Maybe blame fiction? The narrative is that bullies are Other, resembling Moe from Calvin + Hobbes. Embryonic Law & Order villain characters. In reality for bullying to work it has to come from folks the victim respects. The smartest, most popular students are the ones most able to bully.

      • AJD says:

        In reality for bullying to work it has to come from folks the victim respects.

        This certainly doesn’t ring true for me. I didn’t respect any of the people who bullied me in elementary school.

  3. Dan says:

    “I think the proper way to do this study would have been to do an anti-bullying intervention at a couple schools, leave a couple similar schools as controls, and if the anti-bullying intervention successfully decreases bullying, compare outcomes for children at the two schools. I understand this probably would be logistically impossible, plus you’d have to wait another forty years.”

    This would require finding an anti-bullying intervention that actually works at reducing bullying, with a large enough effect size to have the statistical power needed to find results, and which is unlikely to have effects on life outcomes that operate via pathways other than reduced bullying. Quite the logistical challenge. It would be easier to just have an intervention that increases bullying.

    • zslastman says:

      E.g. have a popular TV program feature an episode in which some random minority, say those of a particular hair color, are mocked and portrayed as low status victims OH WAIT MATT AND TREY PARKER ALREADY DID IT FOR US FUCK YOU VERY MUCH.

    • Deiseach says:

      Plus that only works in school. Outside school, when the pack (and it’s never a single bully, it’s him/her and their coterie) are lying in wait to beat up Joe or Sally who tattled on them to the teacher, or who live on the same street or who can meet their victim on the playground, in the shops or elsewhere, school intervention does very little to prevent the bullying going on.

  4. Temporarily Anonymous says:

    I’m not posting this with my usual handle, because it’s going to get uncomfortably personal for me.

    I really want to talk about an implication you have here that really disturbed me. At one point in the article, you say “Those must be some bullies,” apparently implying that you don’t think bullies can have the sort of lifelong effect described. I do not know if this is true universally, or even often, but I can tell you that I’m quite certain it’s true for me.

    It’s actually a little interesting that I would not qualify for the study here, since I was not bullied at age 7. I wasn’t actually bullied at all in elementary school. I wouldn’t say I was super popular, but I was friends with popular kids, and generally well-liked. Unfortunately, the way our school system worked, I ended up separated from all my friends and even most of the people I knew when I went to middle school.

    This wouldn’t have been bad by itself, but right before I entered school, a water line at our house burst. Rather than use a neighbor’s shower every day, I decided I would shower after gym class, which meant getting naked and showering right in the changing room, in front of everyone. In the very beginning, I wasn’t the only one to do this, but everyone who did so was an immediate target for mocking. The other showering kids stopped after just a few days, but I had a truly incredible stubborn streak when I was younger, and decided that if they were going to mock me, I was going to keep doing it just to spite them. I showered after gym every day all year in 6th grade.

    This ended up just painting a bulls-eye on my back. I was the primary target for all the bullies that year (and I learned later that we had an unusually high number of ‘troubled’ kids, as the administration labeled them). Since I had been separated from most everyone I knew, I didn’t have any support from existing friends, and it was pretty much impossible to make friends while being such a target. The bullies were in most of the same classes as me, so there was no escaping them. Sixth grade was easily the worst year of my life.

    I was not clinically diagnosed, but based off of everything I know I think it’s safe to say that I was severely depressed in sixth grade. I can say for certain it the was first time I had suicidal ideation. I can remember fantasizing about finding a tall building (or tree, since I lived in the woods) and jumping off almost every day. The only thing that stopped me is that I was quite religious as a kid, and firmly believed I would go to Hell if I did so.

    That year had profound effects on me, effects that I can still see. The depression that started that year has returned periodically throughout my life. At various points, I almost failed a semester of college, truly wanted to kill myself, and did drop out of grad school because of the depression.

    Before middle school, I was a fit, thin kid and never had any weight issues. I gained a lot of weight in sixth grade, and have struggled with obesity ever since. This may have been inevitable, since my father is morbidly obese, but it started in sixth grade, and none of my siblings have the same weight problems I do, so I don’t think so.

    My self-esteem and self-confidence was completely ruined in sixth grade. It wasn’t until my senior year of high school that I was able to even start making friends again. At one point, I was actually asked out to a dance by a girl (notably, this girl didn’t go to my school). This was a girl I liked, and had wanted to ask out myself. I wasn’t able to say yes, simply because I couldn’t imagine her enjoying herself, and didn’t want to ruin the casual friendship that we had. This was in my senior year of high school. For some contrast, in 5th grade, before all this happened, I had actually asked the girl I had a crush on at the time out on a date.

    I am almost 30 years old now. In my whole life, I’ve only had two girlfriends, and in neither case did we get together because of my own initiative. I still can’t really make friends, because I find it impossible to ask if I can hang out with people. As far as I can tell, I’m broken, and always will be, and it certainly seems to me that I can trace it all back to being bullied in sixth grade.

    Hopefully you can understand why it disturbs me for you to apparently flippantly dismiss the long-term effects of being bullied.

    • mareofnight says:

      To add another point to the sample, someone I know well had a similar experience. Because of that, it seemed reasonable to me that a large part of it could be causation.

    • Scott Alexander says:

      I don’t think my expression of skepticism at the very high effects seen in this study is equivalent to flippancy about being bullied.

      Note that 28% of the children in the study were moderately bullied and 15% were frequently bullied. I don’t think 15% of children go through what you went through. Even if they did, I don’t trust parents to report it accurately enough to detect because of the reliability issues mentioned at the bottom.

      I’m sorry about your experience. I have only been bullied a little, but it was enough to know it sucks.

  5. Michael Edward Vassar says:

    The .11 inter-rater reliability is the real cover story here. WOW! I mean, WOW!

  6. Ilya Shpitser says:

    This is great — now I don’t need to set up my own blog :). Keep up the good work, Scott!

  7. US says:

    For some reason the post reminded me of this..

    I take the incredibly low inter-rater agreement to be the main story here, or at least the part I’d like to know more about. Is the correct interpretation here that the parents have almost no clue about whether their children are bullied or not, while the teachers do, or is it rather that the parents and the children/teachers have a different interpretation of the same data? If the teachers have a much better idea than the parents, then why do they not tell the parents? If they do, what’s going on?

    Lots of questions here, but ‘I wonder if we dealt with all the confounders’ probably isn’t one of them…

    • nydwracu says:

      Children learn very quickly that their parents will side with administrations and authority figures against them. Children also learn very quickly that administrations and authority figures will side with bullies against them.

      • ozymandias says:

        Alternately, I never told my parents when I was being bullied because I had figured out that they would get upset, want to solve the problem, and decide to solve it by lecturing *me* about all the things *I* was doing wrong. Or, in particular in the case of sexual harassment, to tell me they had a crush on me because I was beautiful and I should be less sensitive. >.> I suspect many parents are similarly useless and this is known to their children.

      • Army1987 says:

        their parents will side with administrations and authority figures against them

        I think that’s no longer the case — or, at least, so goes the stereotype of present-day parents (and listening to teachers I don’t get the impression that it’s that off).

      • US says:

        If parents systematically side with the teachers, the cross-informant agreement between parents and teachers would be high – and if the overall agreement is around 0.1 this is unlikely to be the case (however the abstract of the Rønning et al paper doesn’t tell you much and I haven’t gone looking for a copy of the paper – perhaps the teacher-parent agreement is higher than the others).

        As for whether parents will side with administrators this is not really something I know anything about, but it seems to me that there are likely to be some heterogeneities at play here which are important. Say you have two types of parents, ‘smart successful parents’ (-‘SSP’) and ‘not very smart and not very successful parents’ (-‘nSSP’). Which type of parent is more likely to side with the administration, and which type of parent is more likely to have children who’re bullied? If ‘you get bullied because of low status’ and low status is ‘hereditary’ (bullied children are likely to have nSSP), parents of bullied children may still be quite likely to side with the administration, despite the fact that culturally a lot of SSP no longer do this, at least not to the extent they used to. Maybe they stratified parents according to SES groups in the Rønning paper and I’m all wrong, but this is at least an idea. I’m aware that lots of smart children with successful parents get bullied as well. However we get back to the point here that If parents of bullied children are likely to side with the teachers/administration then we’d expect high inter-rater agreement.

        @ozymandias: This is a good point which I probably should have included as well. The ‘parents are useless’ angle is not unfamiliar to me and it certainly seems likely to at least be part of the explanation.

        • US says:

          Aha, the link provided by Anonymous (thanks!) did allow me to have a look at the Rønning paper. They don’t stratify parents according to SES or anything like that.

          One relevant passage:

          “There was very low cross-informant agreement about bullying and victimization of eight-year-old boys. As expected, teachers reported a particularly high level of ‘‘frequent bullying’’ (6.6%) compared to the children themselves (2%) and parents (1%). That teachers report such a high percentage of frequent bullying compared to the children themselves is interesting, in view of the fact that incidents of aggression and bullying in the school are often subtle, indirect, and not easily observable by teachers [24]. On the other hand, the difference in proportion between teacher and child concerning bullying could mean that children’s lower report is a reflection of bullying being a socially undesirable act. Children identified as aggressors may be less likely than others to identify aggressive acts as bullying [20]. Thus, these individuals may downplay or justify their bullying behaviour by not identifying it as bullying [6]. […] As expected, children themselves reported being victimized (6.6%) more often than parents (1.8%) or teachers (2.3%). The results suggest that victimization is not always recognized by parents or teachers. […] Several studies have uncovered that, across cultures, about 50% of the children who were bullied did not tell parents or teachers [13, 36].”

          Parents and children don’t like to think of themselves (/their children) as bullies, whereas the teachers are less hesitant about categorizing the children as such. Parents also seem not to like thinking that their own children are being bullied. And/so a lot of bullying is not reported (because ‘parents are useless’ and/or perhaps because the children think they’ll just side with the administration?).

          Hmm… Just realized that my last part of the previous comment: “we get back to the point here that If parents of bullied children are likely to side with the teachers/administration then we’d expect high inter-rater agreement” may arguably be incorrect, depending on how important this factor is in deciding whether children tell their parents about bullying experiences. The idea was that if parents and teachers coordinate then their ratings will be similar. However if children make such coordination very difficult on account of not telling their parents about being bullied (or about engaging in the bullying of others) because of a fear that the parents will just side with the teachers (which may be problematic e.g. but certainly not only because “many teachers tend to exclude many indirect forms of bullying, such as spreading rumours, intimidations and social exclusion”), then the ‘expected coherence’ between the teacher and parent ratings may be lower than would otherwise be the case.

        • nydwracu says:

          Just realized that my last part of the previous comment: “we get back to the point here that If parents of bullied children are likely to side with the teachers/administration then we’d expect high inter-rater agreement” may arguably be incorrect, depending on how important this factor is in deciding whether children tell their parents about bullying experiences.

          Right, that’s what I meant: if telling their parents doesn’t make their situation better (and especially if it makes it worse, which is my guess), they just won’t do it.

  8. Anonymous says:

    The author isn’t Reming, but Rønning (or Ronning, if you prefer). And your link is broken, but I cannot post links today, so I cannot provide the correct link, at libgen. — dk

  9. Daniel Speyer says:

    How you actually do the study: find something we know causes being bullied but is unlikely to directly affect life outcomes (eg moving towns in a critical year, red hair but only after South park…) and see how it correlates to life outcomes. If bullying is causal, we’ll see a strong correlation (two step cause chain) but not if mutually caused.

    • Desertopa says:

      Well, probably not a *strong* correlation, because the correlation those factors have with bullying is itself not strong enough. There are some characteristics which will pretty reliably result in kids being bullied, and there are some characteristics which probably do not directly correlate to social problems in adulthood which can result in kids being bullied, but I can’t think of any characteristics which are unlikely to directly correlate to social problems in adulthood which would *reliably* result in kids being bullied.

    • Scott Alexander says:

      I think switching schools has been found to correlate with lots of bad outcomes, but I think people usually think of that in terms of lost popularity and social self-confidence, which is different than bullying directly.

    • Ilya Shpitser says:

      You also need to make sure this something isn’t itself affected by the confounders you are trying to deal with (in which case the something you are describing has a name — “instrumental variable.”)

  10. Troy says:

    Most of the findings of this study were in the “frequently bullied” population, and part of the criteria for “frequently” was bullying both at age 7 and age 11. Unless that’s just one really persistent guy, that means the child has gotten independently selected for targeting in two different environments. That could be bad luck but could also be the effect of high inter-bully reliability in what (persistent) qualities make a good victim.

    There is another possibility: it could be that getting bullied initially has bad effects on a child that make him more susceptible to bullying in the future. Presumably this is what the hypothesis of the study would predict. I share your skepticism that bullying always has such lasting bad effects, but it does seem plausible that in at least some cases a kid who is seriously bullied in grade 6 might lose self-esteem, lose friends, etc., thus making him a more attractive target for bullies in grade 8. (On the other hand, perhaps earlier bullying makes some kids harder to bully, as they learn how to react to it. I think that was the case with me — I was a very sensitive child and found it hard to keep from crying when insulted or bullied. Bullying in elementary school taught me that I needed to control my emotions, which I think may have helped me avoid it later. Of course there were bad effects of the bullying too, so in the end it might have been a wash.)

    Also, it’s not clear to me that all traits that make children more susceptible to bullying would also make them less successful later in life. For example, I take it that finding learning fun makes one more susceptible to bullying in at least some environments, but also tends to make one more successful later in life. (Indeed, this is one reason I find these study results so surprising — I’d have thought that the nerds like me who were picked on in school are more successful now, because we’re smarter.)

  11. Deiseach says:

    Most of the findings of this study were in the “frequently bullied” population, and part of the criteria for “frequently” was bullying both at age 7 and age 11. Unless that’s just one really persistent guy, that means the child has gotten independently selected for targeting in two different environments.

    You’re not familiar with the British school environment. This is a study that started in the 1950s. The reason they picked age 11? That’s the year pupils (used to) do the 11 plus, that is, the exams at the end of primary school which sorted out who went on to grammar school and who remained in secondary modern/comprehensives.

    7-11 year olds in Britain and Ireland would mainly (it depends; schools used to be segregated by sex so after age 7 the boys would go to all-boys schools and the girls to all-girls schools where before age 7 the classes were mixed) be all in the one primary school. So if the bully is in your class, or the year ahead of you, as you move through the school, the bully is still there. So you’ve got a good chance of being consistently and continuously bullied by the same person and his or her gang between those ages.

    After the age of 11, you may split up and go to different schools, so that’s why that was the cut-off point.

    • Scott Alexander says:

      Thanks.

    • Slow Learner says:

      I have just read the comments wanting to make that very point. I read the post and thought “But of course it’s logical, that the child at the bottom of the social order in primary school is still at the bottom LATER in primary school.”
      Not to mention, most primary schools are relatively small, so the number of potential victims is limited; and all the shake-ups of adolescence which will shift dominance hierarchies (I was a small child right up until age 14 or so when I became a lot taller and a bit broader) are in the future.

      Whether the scale of the effects is plausible is another question, but the dynamics of bullying in 7-11 year olds are entirely so.

  12. tut says:

    Do you remember when you wrote about the marshmallow studies. A certain test in small children predicted all manners of things about those children after they grew up. Of course conservatives crowed a bunch about that.

    Then somebody did a modified experiment with the same test, but this time they manipulated how much the kid trusted the adult giving the test. And it turned out that a much bigger proportion of the children passed the test in the high trust condition.

    You wrote about that as a debunking of the original findings. Of course a kid’s trust in adults controls all of these important outcomes many decades later! The marshmallow test has nothing to do with internal qualities of people!

    Now let’s get back to the study you talk about today. Specifically let’s talk about the part where the inter rater reliability between children and their parents is 0.11. That means that parents don’t know shit about whether or not their children are being bullied.

    So how do parents not know shit if their children, whom they love more than anything else, are being tortured daily for years? Because the children don’t tell them. Bullied kids don’t tell their parents, and they don’t tell adults in school who would tell the parents, because they have learned that confiding in adults leads to getting hurt.

    If you find it natural that experiencing broken trust can explain all manners of outcomes decades later, then I don’t see why you would think it so extraordinary that broken trust plus regular beatings and shaming plus severely reduced opportunities to have normal relationships with ones peers would have long term consequences.

    • tut says:

      One more thing: The way you talk about anti bullying programs means that you might not find it obvious how you get hurt when you talk to adults about bullying. My first reaction, before I would have been able to express anything about what happened was “I don’t want to go to school” and “I don’t want to … [get into this situation in school where I know that I will get hurt]”. Their reaction was of course always “You must” followed by guarding me extra well until they had made sure I wouldn’t escape. If you are socially smart that’ll probably be enough to teach you that adults are not on your side. I was not so quick on the uptake, so I have some experience of anti bullying programs.

      All schools have anti bullying programs. Not having them, or not training teachers in them, or not following them whenever there is a suspicion of bullying would be a legal liability.

      Here is how a program works. Details vary, but essentially an anti bullying program is an overly complicated protocol for informing your bully that you have snitched. After that you of course get a beating. If you answer truthfully, or just don’t come up with a good enough explanation when your parents ask about your contusions or bloody clothes the bully is informed that you have snitched again. Rinse repeat until you have learned not to snitch. Problem solved.

  13. P says:

    This twin study found that both being a bully and being bullied have high heritability (61% and 73%, respectively).

  14. Polymath says:

    I see it pretty simply.

    (1) Bullies exist and can do horrible damage to kids when the adults surrounding the kids are incompetent or insensitive.
    (2) When the adults surrounding kids are competent and sensitive, the damage done by bullying and the incidence of bullying can be reduced to a very low level.
    (3) Most adults in most public schools in the USA are incompetent or insensitive or both.

    (Sometimes this is because the system’s incentives make or select for incompetent or insensitive teachers; I count a teacher who is afraid to help a bullied kid because of bad school district policies as “incompetent” because if the teacher is sensitive enough to understand who is bullying whom and how, a great deal can be accomplished, even if it has to be outside the system, such as giving the bullied kid a lot of emotional support and encouraging his parents to fight for him or switch him to a better school).

    What is there to say that these three points don’t cover?

    I have always hated bullying more than practically anything else, and it was always obvious to me that the kids who were bullies included the ones who would grow up to be criminals, and although there were lots of bullying attempts I was so convinced of my own righteousness that I never stopped insisting that my schools stop tolerating behavior that would be unacceptable if an adult did it to another adult (thus, hitting me or taking my stuff was absolutely and utterly unacceptable but I didn’t think the school needed to stop people calling me names). The one school I went to that wouldn’t act properly when this happened, I simply refused to attend (I was in 8th grade).