Welcome to the Science Fictions newsletter. Or, if you’re one of my previous subscribers from Substack – welcome to the newsletter’s new home. Science Fictions is continuing here at i, and will now come out every Thursday. I’ll keep covering controversies in science, picking apart bad research papers, and digging into the data on interesting scientific questions.
As ever, I’d love for you to get in touch with any stories of bad science, or anything you’d like me to look into – my contact details are at the bottom of this page.
A few years ago, I spoke in a debate entitled “Parenting Doesn’t Matter – Or Not as Much as You Think”. I was invited because I’m a psychologist by training, and I have a particular interest in the nature vs nurture question: what makes people turn out the way they do?
I was speaking for the motion, arguing that science shows that the impact of parenting on children’s life outcomes is a lot smaller than people think (just for the record, our side lost).
After the debate, one of the opposing team asked me if I had any children. I said no, and she nodded sagely. “You’ll see,” she said, in what I felt was a somewhat unfair comment given that the debate was supposed to be about hard evidence.
Well, that’s about to change. My wife is in the final days of her pregnancy, and by the time you read this we might already have our first baby (it’s a girl). And despite arguing in the debate that parenting effects are overrated, I still feel pretty nervous about the whole thing. I thought I’d take this opportunity to re-assess the arguments. So: does parenting matter?
Defining the question
Of course, it’s a very ambiguous question, since both “parenting” and “matter” can be defined in all sorts of ways. Some types of parenting obviously matter: if a child suffers severe neglect or abuse at the hands of their parents, it’s no surprise if it changes the way they turn out. No debate there.
If you include under “parenting” things like poverty, you also won’t find much disagreement that parenting matters. Living in a poor area of town, with perhaps worse nutrition and more pollution, is very likely to cause worse life outcomes. But although poverty is a characteristic of a child’s parents, I don’t think it’s really what people mean when they talk about “parenting”.
And that’s where the question becomes more intriguing. In cases where a child’s basic needs are met, within the normal range of how parents act, does the way they act matter? Which “parenting style” they use? Whether they coddle and spoil, or whether they’re strict and authoritarian? Whether they give unconditional love and constant attention? Whether they’re “securely” or “insecurely” attached to their child? Whether they do “hot-housing” or “helicoptering”, or follow any of the other hundreds of pieces of parenting advice that are flung around every day?
And you might ask: matter for what? The only answer that a scientist can give is: for things we can measure. For your child’s personality: how outgoing, anxious, or friendly they end up being. For how smart they are, or how they do at school. For their risks of developing a mental health problem like depression or schizophrenia, and for how happy they are with their life.
The “obvious” answer
Your first answer might be: of course parenting matters! Parenting looms large in our culture for so many reasons, but in large part due to Sigmund Freud, who put massive emphasis on all the ways your parents can screw you up (and, er, your own deep desires to screw your parents).
Despite the patent absurdity of so many of Freud’s ideas—he once suggested asthma was due to kids overhearing their parents having sex and trying to imitate the heavy breathing—they’re still everywhere. Open nearly any biography or autobiography and you’ll find speculation about the subject’s relationship with their parents and how it set them up for life.
But beyond that, there are many studies you can point to which find correlations between the way parents act and the way their kids turn out. Let’s just look at three randomly-chosen studies that have already been published in 2023:
- A study in the Journal of Criminal Justice found that lower levels of parental attachment were associated with higher levels of crime and deviance in their children when they became adolescents;
- A study in Child & Youth Care Forum found that parents who displayed more “supportive engagement” had kids who did better at school;
- A study in the Journal of Affective Disorders found that, during Covid-19, parents who were more anxious and depressed had kids who were more anxious and depressed.
Parenting effects, right? Wrong. Or at least, not at all assured, for one big reason: genetics.
In these studies, the characteristics and behaviours of parents were being correlated with those of their biological offspring. Take the final study mentioned above: depressed parents tend to have depressed kids. Why might that be? It could be because of the way the parents are acting towards their kids. But it could also be that they’ve passed on DNA that we know predisposes people to depression: just like taller parents have taller kids on average, depression-proneness also runs in families for genetic reasons.
These studies, which are purely correlational and don’t look at any genetic data, can’t tell the difference between effects due to parenting and those due to genetics. But the assumption is almost always that it must be due to something the parents are doing.
I’m not saying all these associations between parenting and child outcomes were definitely due to genetics. They might not have been – they might have been 100 per cent due to parenting! But using this kind of study, we simply can’t know. Indeed, doing this kind of research is a waste of time unless you do something to account for genetic relatedness.
It’s not like this important argument was considered and then dismissed for good reason: the word “genetic” isn’t mentioned once in any of the three studies above. It’s just bad science, and could be highly misleading if we tried to work out which specific parental behaviours matter the most for their kids’ development.
The actual answer
So what about when scientists do take genetics into account? These are usually studies of twins, both identical and fraternal, which give you three quantities:
- Heritability: how much the differences in a given trait are related to differences in genetics;
- Nonshared environment: how much the differences in a given trait are related to unique things that happen to you, but not your twin;
- Shared environment: how much the differences in a given trait are related to things that happen to you and your twin – including, presumably, parenting.
In general, for the sorts of things we discussed above—personality traits, vulnerability to mental illness, how well you do on an IQ test—the shared environment effect is very small, or is moderate in early life and then peters out. Put another way, most of the reasons people resemble their parents, at least on these kinds of traits, in the long run, are to do with nature rather than nurture.
There is a big exception, though: educational attainment. A large twin study from 2021 showed that the shared environment might explain about 30 per cent of the overall differences in schooling, even in adulthood. Could it be due to something wholesome, like parents helping their kids to study for exams, leading them on to better qualifications? Or to parents using their connections to get their kids into higher levels of education? Or something else? We don’t really know.
Education aside, though, there’s little evidence from properly-controlled research for a strong influence of parenting. The parents of the people in these studies will have varied in all sorts of ways: warmth, interaction, attachment, encouragement, goal-setting, conscientiousness, and so on. And yet those variations have left at best a small impression on the children’s traits.
This becomes more obvious if you think about siblings: if parenting had a strong effect, why are children from the same family, with the same parents, often so different from one another? It seems like parents know about this. As one 2019 study found, if you ask people how genetic various traits are, and line up their guesses with measured heritability from studies, accurate guesses are made by mothers with more than one child.
The first counter-argument to all the above might be that if you look at deliberate parenting interventions—where researchers have done experimental studies with a randomised group of parents being taught a specific skill, following them up some time later—there’s abundant evidence for parenting effects. There’s a helpful meta-analysis from 2021 that gathered together 102 studies from all over the world: on attachment, language development, parent-child interaction, and more, all in the first three years of life. They concluded that parenting interventions were indeed effective at improving early child development.
It seems plausible to me that there are better ways of doing lots of parenting tasks. But I would urge caution about the meta-analysis. The studies varied dramatically, meaning that the average effects were from a hotchpotch of different designs, sometimes with just a few dozen participants. It’s also near-impossible to blind a study of that nature – to hide whether you’re in the intervention or the control group to guard against expectation effects (indeed, every single study they looked at was at high risk of bias for this reason). It’s also not clear how long-term the effects are.
Another response might be that this is all just a straw man argument. Nobody really believes they can mould their kids’ deepest personality traits. When people talk about parenting mattering, they’re simply not talking about the stuff that psychologists are measuring.
Maybe. But I’d ask anyone making this argument to tell it to the purveyors (and buyers) of parenting advice and trends that fill up social media, books, and TV shows – it’s doubtful that people would part with so much money to learn about “sittervising” or “unschooling” or “indigo children” or any number of other unevidenced fads if they didn’t believe it would make much difference.
They should also tell it to the innumerable parents who expend huge efforts in pushing their children to do this or that activity, to become this or that kind of person – then spend the rest of their time worrying themselves sick that either they’re being a bad parent, or that the child isn’t living up to expectations.
Parenting does matter – just not the way you think
I remember there being a poster in my high school music department that announced that music lessons “make children more intelligent”. This is the wrong way to think about it. Music lessons aren’t valuable in this very “instrumental” (no pun intended) way, where you make your kids do them because they might gain a few IQ points. They’re valuable because learning about music is one of the best things a human can do.
Interests. Tastes. Manners. Languages. Traditions. Cultural experiences. None of these are particularly measurable, but they’re all things we can do with, or encourage in, our children. Aiming to fill our children’s lives with nice things just because they’re nice things is a much healthier outlook on parenting than one where we fret constantly about whether we’ve helped them meet this or that goal.
As Steven Pinker put it in his book The Blank Slate:
“There are well-functioning adults who still shake with rage when recounting the cruelties their parents inflicted on them as children. There are others who moisten up in private moments when recalling a kindness or sacrifice made for their happiness… If for no other reason, parents should treat their children well to allow them to grow up with such memories.”
It used to be the case that parents’ behaviour was blamed for kids getting psychiatric disorders like schizophrenia, or developmental conditions like autism. We now know that these conditions are largely due to genetics, and parents are off the hook.
Isn’t that a relief? For the most part, your children’s personalities, and many of their other characteristics, would’ve been the same regardless of your “parenting style”. If you think about parenting this way, you can relax, focus on enjoying spending time with your children and creating those happy memories, and watching them grow into the people they were likely always going to be.
We shouldn’t, of course, choose to believe or disbelieve a scientific finding because it feels nice to do so. But regardless, the fact that parenting effects are generally small should be a relief for many who worry that the way they act during their son or daughter’s childhood could leave an indelible mark.
And although at first it might make you feel like you have no control over how your child will turn out, the question you should ask is whether you really wanted to have control over another human being in the first place.
That’s what I’m telling myself, anyway. Wish me luck.
Other things I’ve written recently
- You might’ve seen a lot of claims about how scientists have reversed ageing in mice. I cast a bit of doubt on that idea in an article earlier this week.
- There is some good news, though: the weight-loss drug semaglutide continues to show results, and could be the thing that helps millions of people get control of their weight.
- The Doomsday Clock is very annoying and cringe. It’s also unscientific. Here’s why.
Science link of the week
I know I mentioned Saloni Dattani last week. But you’ll forgive me for mentioning her again when you read her new, enormous Substack post on all the exciting science stories you’ve missed since late 2022.
Thanks for reading the Science Fictions newsletter. If you’ve any ideas for other questionable scientific claims I should be looking into, I’d love to hear from you: firstname.lastname@example.org or tweet me @StuartJRitchie. If you’d like to get this newsletter direct to your inbox, every single week, you can sign up here.