I started by writing about the basic demographics and politics of my readers. Then we talked about your likes and dislikes, and the things you’re wrong about. To wrap up the Richard Hanania newsletter series, I’m going to talk about birth order effects. I didn’t think this was something that actually existed. Many years ago I read Judith Rich Harris’
I would be the youngest of 5 in my *birth family*, but the only one of the 5 to read Hanania, or read at all. There is a substantial IQ / temperament gap between me and those siblings.
In my adopted family - adopted at a coupla months, I am the oldest of two, where again I'd be the only reader and beneficiary of a sizeable IQ advantage.
If there is a reason (I'm an only child and have no dog in the fight) it might be that older siblings grow up more accustomed to (and liking) having their opinions listened to. Even if the site itself does not provide a way of voicing opinions, the person might still be attracted a "preparation" for opining in other fora.
Right. Could also be that older siblings are more likely to respond to surveys (for the reason you suggest—liking to express their opinions—or for some other reason).
I thought the book was pseudo-science. No studies, just the author's personal observations. (Cards on the table: In a different book Lehman attacks the Behavior Genetics of Judith Harris, and I probably hate him for being closer to the "Blank Slate" side of the spectrum.)
While Lehman nailed my personality in his Birth Order book, I assumed this was just the Barnum effect; similar to the "5 Love Languages"--a book that's not unreasonable, but filled with zero studies, controlled experiments or factor analysis--but that probably does improve people's lives.
That said, I find it interesting that I'm such an outlier! I'm the youngest of two, but I'm definitely the smartest member of my family (in terms of education, income and job placement). I would be fascinated to see what the Big Five traits of your readership are! (That would be hard to gather data on, though.)
Finally, I hate to rain on the parade in the last paragraph, but--when I read it the first thing I thought was, "Wow. Maybe some people are much more likely to change than others," rather than assuming that you were the catalyst. I--of course!--deeply admire you, your accomplishments and your substack. While your work has certainly changed my way of thinking, I'm not sure if your views on Civil Rights or Women's tears would cause me to parent differently.
P.S. I was excited to hear your interview on the Tom Woods show! (see here: https://www.youtube.com/watch?v=4R58TI1CanI) Tom didn't mention me, (see the 1:22--1:38 mark in the video) but I was the one who handed him your name at a live event in Florida about two weeks before you were on the show. Really hope you sync up with him again. Going to promote your work in our Private chat!
The key book on birth order was written 25 years ago. It's "Born to Rebel," by Frank Sulloway. He's one of those odd ducks you'd occasionally see in the bowels of your college library, mid-30's, working on a magnum opus, but not teaching many classes (if any). His t-shirt was stained with sweat and he surrounded himself with 150 biographies of scientists through the ages, carefully taking notes, neglecting to eat more than one meal a day. He figured it out down there.
I'll check that out, thanks. I'm probably fairly difficult to move out of his camp, though, because his statistics were very convincing. If she took him to task for having a Nurture vs Nature position, I'm not sure she read him correctly. His was an evo-psych treatise: Younger siblings have to do *something* to get resources, so they improvise.
Correlate with conservatism (his book argues that firstborns and onlies are more conservative after all). Hanania's (you want a Dr.? you've said you hated academia but I want to respect my host) already pointed out his fans lean right. If firstborns are conservative, and Hanania fans are conservative, then it's quite possibly the effect is explained by that.
“His methodology was straightforward. If there were no birth order effects, then whenever there is a reader who grew up in a two-child family, there should be a 50% chance that they were a firstborn.”
Wait, hang on — this seems wrong.
Imagine a town where every family has two children. This is a funny town, and in this town the families all decided to have the second child last year.
In this town, the chance that a reader of Scott Alexander is the elder child of two is close to 100%.
That’s an artificial example to show that the argument — at least as presented — can’t work. It depends on the age distribution but generally you expect bias.
More realistically, you expect bias toward being the eldest child since the probability of encountering/reading Scott generally goes up in the age range zero to, say, 20. If you asked, instead “what are birth order effects for wearing diapers”, using this method, you’d find the opposite—why are all these second-borns so immature?
Well if you look at the chart taken from Scott’s blog, you would with that theory expect a drop off with each increment, being a third child versus fourth child, etc. That doesn’t look to be driving it, it’s mostly a firstborn effect. It’s harder to tell with my data because sample sizes are smaller.
You’d have to model the age distribution to see. There are *lots* of potential confounders, including via the parents. For example: say your readers are all 25. Some have an older brother at 30, the others a younger brother at 20. This generally would imply that the younger sibling has older parents, and that may affect what he is interested in. That leads to an asymmetry that isn’t causally related to birth order (it’s the parent, not the sibling, driving the effect) but looks like it is.
It’s not that birth order effects aren’t necessarily real, just that the pure fact of the imbalances you’re seeing here are not evidence for or against. They’re something you would expect under pretty generic conditions when birth order effects were minimal.
It’s certainly interesting -- and yes, I think you have stronger evidence when you’re restricting to pairs which are likely to both be of reading age.
If you’re interested in birth order effects, one thing you might check for in your sample is a common result that homosexuality in men correlates with birth order -- but in the other direction (of a pair, the younger is more likely than the older to be gay). Do you see this? The effect is known for older brothers, but you can just assume (as a first guess) that the other sibling is randomly male/female.
It should be measurable in that (considering just your two sibling cases) the raw effect sizes are around 40% for one older brother) -- so 20% since you don’t know older sibling sex. With a male-dominated sample of ~400 “children with one sibling”, it might just be detectable at p < 0.05 (consider this post your preregistration: compute fraction of men in two-child households who are gay, split by birth order, standard effect should be +20% for the second child.)
I don't think your example applies to the real world. As written, it demonstrates that it is *possible* to construct a scenario where second siblings are on average younger than first siblings in some cohort, but it does not demonstrate that we should expect any given real-world sample to have this trait.
There's no reason to believe that second siblings should be on average younger than first siblings.
If I pick randomly 1000 25-year olds who all have one sibling, you would expect to get 500 older siblings and 500 younger siblings. The fact that second siblings are born later would manifest as the second siblings having on average older parents.
"It’s unlikely that age alone is driving these results. In sibships of two, older siblings on average were only about one year older than younger siblings. That can’t explain why one group reads this blog so much more often than the other."
In general, the variance of age within sibships is small compared to the variance of age in Scott's readership, so age cannot be that relevant.
That’s interesting and the very narrow age gap is quite interesting in an of itself. Doctors usually counsel leaving 18 months between children. It might suggest a mechanism.
At less than 12 (maybe 18) months between pregnancy, maternal nutrition becomes an issue for the second pregnancy. It would be interesting to look at age gap effects -- under this hypothesis, it would be more likely to be the elder child when the younger sibling is <12 months.
The mechanism in this case is that the elder child in the pair requires significantly less attention and will have significantly better outcomes than the younger. This might well predispose the elder child to self-educate more. You’d need birthdates to see this effect since it’s pretty tight.
(My knowledge of birth order effects is from studies of scientific ability and that it’s often the *younger* child who is more exploratory, playful, etc -- I.e., the preference for elder in substack self-education would go the other way.)
Your final comment here about scientific ability in *younger* children sounds like the thesis of Born to Rebel by Sulloway. (See my other comment about birth order.) Essentially, his thesis could be used to understand why Hanania Appreciators (HA!) would skew heavily towards eldest children: They help parents w/younger sibs, appreciate "conserving" family resources, commonly appear more "serious" to outsiders, etc. ALL of these aspects of Eldest vs. Younger children - aspects commonly understood by nearly everyone on earth, but not addressed statistically til Sulloway, to my knowledge - could be at play with HAs. We're soberly discussing evo-psych texts, investing in the expansion of our minds for the betterment of our families, while our irresponsible, dissolute, somewhat creative youngest sibs are preening for attention at Occupy Wall Street protests and begging for debt forgiveness for their bachelors degree in 15th Century Indonesian Pottery. It all syncs up.
Birth order effect for intelligence just seems to be real.
I haven't checked if this is true, but interestingly Arthur Jensen wrote that "The disadvantage of being later-born, however, is very slight and shows up conspicuously only in the extreme upper tail of the distribution of achievements."
It should also be noted that a birth order effect could be (in part) a biological phenomenon i.e. the accumulation of harmful mutations in the gametes as parents age.
With a large enough sample size, even an effect on a small portion can be statistically significant.
You don't just have to rely on Judith Harris' empirical work. There's also a theoretical basis in Trivers' theory of genetic conflict: it is not to the advantage of a child to be so moldable by their parents because the child doesn't have identical Darwinian interests with said parents.
Just because you can influence adults doesn't mean you will be good at influencing children(yours or anybody else's). If you send kids to public school, it's almost a guarantee that their peers and teachers will have more influence than you do. It doesn't mean your influence in none, but the probable reason studies don't find much parental influence is because just about everyone has the same upbringing: public school(and most private schools aren't much different).
There's also the well-documented fraternal birth order effect (the more older brothers a man has, the more likely he is to be homosexual). So in addition to a slightly higher IQ or curiosity, firstborns are slightly less gay, which would certainly make them more likely to be Hanania appreciators.
Well done - many thanks. Thoughtless acceptance and application of statistical indicators has become my pet peeve. Whatever happened to relying on one’s own experiences? It’s as if ‘average’ = universal. Ugh.
As I recall, there is an advantage of 1-2 IQ points for firstborns in the intelligence literature (I say this as a second child). I believe it's attributed to greater parental time investment in the first child - they get 100% of the parenting attention until their sibling(s) is/are born and often an equal or greater share after that.
I always thought the most straightforward explanation was average intellectual age of interlocutors.
A first child has a lot of time where their main social partners are adults. In comparison a second child has a wider range where the main social partner is another child.
That seems like a plausible explanation, too. I haven't done a deep look at that research. Anyway, my main point was that there's at least some evidence for an IQ advantage (albeit a small one) for first children .
Really interested whether this is true for women as well. Is there a way to get some sophisticated feminist website to do a survey of their readers and include this question?
Also, are there hobbies or interests that show the opposite effect with first-borns significantly less likely than expected? Or do you just find either no/tiny differences or significant ones in favor of first-borns?
I'd be a major outlier in your study.
I would be the youngest of 5 in my *birth family*, but the only one of the 5 to read Hanania, or read at all. There is a substantial IQ / temperament gap between me and those siblings.
In my adopted family - adopted at a coupla months, I am the oldest of two, where again I'd be the only reader and beneficiary of a sizeable IQ advantage.
If there is a reason (I'm an only child and have no dog in the fight) it might be that older siblings grow up more accustomed to (and liking) having their opinions listened to. Even if the site itself does not provide a way of voicing opinions, the person might still be attracted a "preparation" for opining in other fora.
Right. Could also be that older siblings are more likely to respond to surveys (for the reason you suggest—liking to express their opinions—or for some other reason).
I bet this is older brothers hitting the younger ones in the head.
Or younger kids being less supervised and falling on their heads more.
This is interesting. I read a book called "The Birth Order Book: Why you are the way you are" by psychologist Kevin Lehman. (See here: https://www.amazon.com/Birth-Order-Book-Why-You/dp/0800723848/ref=sr_1_2?crid=FGGVMQUA1H1G&keywords=kevin+lehman&qid=1660913736&s=books&sprefix=kevin+lehman%2Cstripbooks%2C65&sr=1-2)
I thought the book was pseudo-science. No studies, just the author's personal observations. (Cards on the table: In a different book Lehman attacks the Behavior Genetics of Judith Harris, and I probably hate him for being closer to the "Blank Slate" side of the spectrum.)
While Lehman nailed my personality in his Birth Order book, I assumed this was just the Barnum effect; similar to the "5 Love Languages"--a book that's not unreasonable, but filled with zero studies, controlled experiments or factor analysis--but that probably does improve people's lives.
That said, I find it interesting that I'm such an outlier! I'm the youngest of two, but I'm definitely the smartest member of my family (in terms of education, income and job placement). I would be fascinated to see what the Big Five traits of your readership are! (That would be hard to gather data on, though.)
Finally, I hate to rain on the parade in the last paragraph, but--when I read it the first thing I thought was, "Wow. Maybe some people are much more likely to change than others," rather than assuming that you were the catalyst. I--of course!--deeply admire you, your accomplishments and your substack. While your work has certainly changed my way of thinking, I'm not sure if your views on Civil Rights or Women's tears would cause me to parent differently.
P.S. I was excited to hear your interview on the Tom Woods show! (see here: https://www.youtube.com/watch?v=4R58TI1CanI) Tom didn't mention me, (see the 1:22--1:38 mark in the video) but I was the one who handed him your name at a live event in Florida about two weeks before you were on the show. Really hope you sync up with him again. Going to promote your work in our Private chat!
The key book on birth order was written 25 years ago. It's "Born to Rebel," by Frank Sulloway. He's one of those odd ducks you'd occasionally see in the bowels of your college library, mid-30's, working on a magnum opus, but not teaching many classes (if any). His t-shirt was stained with sweat and he surrounded himself with 150 biographies of scientists through the ages, carefully taking notes, neglecting to eat more than one meal a day. He figured it out down there.
https://www.amazon.com/Born-Rebel-Family-Dynamics-Creative/dp/0679758763/ref=asc_df_0679758763?tag=bngsmtphsnus-20&linkCode=df0&hvadid=80333185696759&hvnetw=s&hvqmt=e&hvbmt=be&hvdev=c&hvlocint=&hvlocphy=&hvtargid=pla-4583932713615842&psc=1
See Judith Rich Harris’ No Two Alike for Sulloway’s work. It appears to rest on pretty weak foundations.
I'll check that out, thanks. I'm probably fairly difficult to move out of his camp, though, because his statistics were very convincing. If she took him to task for having a Nurture vs Nature position, I'm not sure she read him correctly. His was an evo-psych treatise: Younger siblings have to do *something* to get resources, so they improvise.
Correlate with conservatism (his book argues that firstborns and onlies are more conservative after all). Hanania's (you want a Dr.? you've said you hated academia but I want to respect my host) already pointed out his fans lean right. If firstborns are conservative, and Hanania fans are conservative, then it's quite possibly the effect is explained by that.
“His methodology was straightforward. If there were no birth order effects, then whenever there is a reader who grew up in a two-child family, there should be a 50% chance that they were a firstborn.”
Wait, hang on — this seems wrong.
Imagine a town where every family has two children. This is a funny town, and in this town the families all decided to have the second child last year.
In this town, the chance that a reader of Scott Alexander is the elder child of two is close to 100%.
That’s an artificial example to show that the argument — at least as presented — can’t work. It depends on the age distribution but generally you expect bias.
More realistically, you expect bias toward being the eldest child since the probability of encountering/reading Scott generally goes up in the age range zero to, say, 20. If you asked, instead “what are birth order effects for wearing diapers”, using this method, you’d find the opposite—why are all these second-borns so immature?
Am I missing something?
Well if you look at the chart taken from Scott’s blog, you would with that theory expect a drop off with each increment, being a third child versus fourth child, etc. That doesn’t look to be driving it, it’s mostly a firstborn effect. It’s harder to tell with my data because sample sizes are smaller.
You’d have to model the age distribution to see. There are *lots* of potential confounders, including via the parents. For example: say your readers are all 25. Some have an older brother at 30, the others a younger brother at 20. This generally would imply that the younger sibling has older parents, and that may affect what he is interested in. That leads to an asymmetry that isn’t causally related to birth order (it’s the parent, not the sibling, driving the effect) but looks like it is.
It’s not that birth order effects aren’t necessarily real, just that the pure fact of the imbalances you’re seeing here are not evidence for or against. They’re something you would expect under pretty generic conditions when birth order effects were minimal.
Well, I just looked at only Boomers in 2-kid families, and they're 71% likely to be first borns. For families of 3, Boomers are 44% likely.
For Gen X in 2 kid families, they're 69% likely to be firstborns. For 3 kid families, they're 51% likely.
Does that satisfy you that there is probably something special about birth order?
It’s certainly interesting -- and yes, I think you have stronger evidence when you’re restricting to pairs which are likely to both be of reading age.
If you’re interested in birth order effects, one thing you might check for in your sample is a common result that homosexuality in men correlates with birth order -- but in the other direction (of a pair, the younger is more likely than the older to be gay). Do you see this? The effect is known for older brothers, but you can just assume (as a first guess) that the other sibling is randomly male/female.
It should be measurable in that (considering just your two sibling cases) the raw effect sizes are around 40% for one older brother) -- so 20% since you don’t know older sibling sex. With a male-dominated sample of ~400 “children with one sibling”, it might just be detectable at p < 0.05 (consider this post your preregistration: compute fraction of men in two-child households who are gay, split by birth order, standard effect should be +20% for the second child.)
https://en.m.wikipedia.org/wiki/Fraternal_birth_order_and_male_sexual_orientation
In this case you’re looking at “probability of sexuality as a function of birth order” so you should have fewer (first order) selection effects.
I don't think your example applies to the real world. As written, it demonstrates that it is *possible* to construct a scenario where second siblings are on average younger than first siblings in some cohort, but it does not demonstrate that we should expect any given real-world sample to have this trait.
There's no reason to believe that second siblings should be on average younger than first siblings.
If I pick randomly 1000 25-year olds who all have one sibling, you would expect to get 500 older siblings and 500 younger siblings. The fact that second siblings are born later would manifest as the second siblings having on average older parents.
Scott wrote:
"It’s unlikely that age alone is driving these results. In sibships of two, older siblings on average were only about one year older than younger siblings. That can’t explain why one group reads this blog so much more often than the other."
In general, the variance of age within sibships is small compared to the variance of age in Scott's readership, so age cannot be that relevant.
That’s interesting and the very narrow age gap is quite interesting in an of itself. Doctors usually counsel leaving 18 months between children. It might suggest a mechanism.
At less than 12 (maybe 18) months between pregnancy, maternal nutrition becomes an issue for the second pregnancy. It would be interesting to look at age gap effects -- under this hypothesis, it would be more likely to be the elder child when the younger sibling is <12 months.
The mechanism in this case is that the elder child in the pair requires significantly less attention and will have significantly better outcomes than the younger. This might well predispose the elder child to self-educate more. You’d need birthdates to see this effect since it’s pretty tight.
(My knowledge of birth order effects is from studies of scientific ability and that it’s often the *younger* child who is more exploratory, playful, etc -- I.e., the preference for elder in substack self-education would go the other way.)
Your final comment here about scientific ability in *younger* children sounds like the thesis of Born to Rebel by Sulloway. (See my other comment about birth order.) Essentially, his thesis could be used to understand why Hanania Appreciators (HA!) would skew heavily towards eldest children: They help parents w/younger sibs, appreciate "conserving" family resources, commonly appear more "serious" to outsiders, etc. ALL of these aspects of Eldest vs. Younger children - aspects commonly understood by nearly everyone on earth, but not addressed statistically til Sulloway, to my knowledge - could be at play with HAs. We're soberly discussing evo-psych texts, investing in the expansion of our minds for the betterment of our families, while our irresponsible, dissolute, somewhat creative youngest sibs are preening for attention at Occupy Wall Street protests and begging for debt forgiveness for their bachelors degree in 15th Century Indonesian Pottery. It all syncs up.
Wouldn't this study by relevant? https://www.pnas.org/doi/abs/10.1073/pnas.1506451112
https://www.pnas.org/doi/10.1073/pnas.1506451112
Birth order effect for intelligence just seems to be real.
I haven't checked if this is true, but interestingly Arthur Jensen wrote that "The disadvantage of being later-born, however, is very slight and shows up conspicuously only in the extreme upper tail of the distribution of achievements."
It should also be noted that a birth order effect could be (in part) a biological phenomenon i.e. the accumulation of harmful mutations in the gametes as parents age.
With a large enough sample size, even an effect on a small portion can be statistically significant.
You don't just have to rely on Judith Harris' empirical work. There's also a theoretical basis in Trivers' theory of genetic conflict: it is not to the advantage of a child to be so moldable by their parents because the child doesn't have identical Darwinian interests with said parents.
Just because you can influence adults doesn't mean you will be good at influencing children(yours or anybody else's). If you send kids to public school, it's almost a guarantee that their peers and teachers will have more influence than you do. It doesn't mean your influence in none, but the probable reason studies don't find much parental influence is because just about everyone has the same upbringing: public school(and most private schools aren't much different).
This might have little to do with nurture.
Wasn't there a study showing firstborns getting more Testo in the womb?
There's also the well-documented fraternal birth order effect (the more older brothers a man has, the more likely he is to be homosexual). So in addition to a slightly higher IQ or curiosity, firstborns are slightly less gay, which would certainly make them more likely to be Hanania appreciators.
First-born gay Hanania appreciator here.
It's good to be an outlier!
Well done - many thanks. Thoughtless acceptance and application of statistical indicators has become my pet peeve. Whatever happened to relying on one’s own experiences? It’s as if ‘average’ = universal. Ugh.
Do you have any ideas about what might explain the birth order effect?
As I recall, there is an advantage of 1-2 IQ points for firstborns in the intelligence literature (I say this as a second child). I believe it's attributed to greater parental time investment in the first child - they get 100% of the parenting attention until their sibling(s) is/are born and often an equal or greater share after that.
A more likely explanation of the effect is increased germline mutations and other biological aging effects in parents.
I always thought the most straightforward explanation was average intellectual age of interlocutors.
A first child has a lot of time where their main social partners are adults. In comparison a second child has a wider range where the main social partner is another child.
That seems like a plausible explanation, too. I haven't done a deep look at that research. Anyway, my main point was that there's at least some evidence for an IQ advantage (albeit a small one) for first children .
Really interested whether this is true for women as well. Is there a way to get some sophisticated feminist website to do a survey of their readers and include this question?
Also, are there hobbies or interests that show the opposite effect with first-borns significantly less likely than expected? Or do you just find either no/tiny differences or significant ones in favor of first-borns?