Why are we not much, much, much better at parenting?

We’ve come a long way, as a species. And we’re better at many things than we ever were before – not just slightly better, but unimaginably, ridiculously better. We’re better at transporting people and objects, we’re better a killing, we’re better at preventing infectious diseases, we’re better at industrial production, agricultural and economic output, we’re better at communications and sharing of information.

But in some areas, we haven’t made such dramatic improvements. And one of those areas is parenting. We’re certainly better parents than our own great-great-grandparents, if we measure by outcomes, but the difference is of degree, not kind. Why is that?

Down to the market…

Let’s turn first to the system that created the desktop computer and the perfected the mobile phone: the relentless churn of global capitalism. Why aren’t companies selling us products to make us super-parents?

But in parenting capitalism, who is the consumer? You could take the kids themselves as the consumers, as they’re the primary concerned. But they have no purchasing power, which means they don’t count. In practice, the parents are the consumers. So the strongest feature of the free-market – that the purchaser directly experiences the quality of the product – is diluted.

And capitalism has indeed responded to parents’ demand, making available a whole variety of products, from car seats to push chairs to nappies to baby monitors. Problem is, though, parents aren’t very good consumers, in that they have a limited ability to figure out (and demand) what they actually need. In part, this is because they have few children apiece: most parents in rich countries (i.e. the most important consumer-parents) will have one, two, or three children, seldom much more. Bearing in mind that children have different personalities and change as they grow, we can see that the average parent will have very little experience in, well, parenting.

This analogy may not seem fair – we wouldn’t say someone didn’t know about televisions or phones, simply because they’d only owned a handful. If they’d interacted with them, customised them extensively and read up on lots of online advice, we’d say they were very knowledgeable consumers! But a child is a much more complicated being than a television or even a phone. Trying to get a child to do something (or, even more challenging, to become something) is much more difficult that following an instruction manual. Moreover, most of the goals of parenting are long term: they want their child to grow up a certain way. But humans are quite bad at estimating the results of different interventions, if the feedback only comes years later. One needs only to see the plethora of different parenting guides and opposed schools of upbringing thought. Such variety couldn’t maintain itself if it were easy for parents to see which methods worked and which didn’t.

Thus parents are poor at knowing what they need, and hence make ineffective consumers from the economic perspective. Companies will cheerfully sell them lots of colourful tools and nicknacks, but actual improvements will be limited.

The visible hand

If the market can’t provide it, what about the government? If governments can save millions of lives through vaccinations, surely they can improve the lot of children and their parents?

And to some extent, they have. By addressing the excesses of child abuse (though removal of the children, or, more importantly, through the threat of removal) and through the provision of universal education, governments have improved the average outcomes for many families.

But those examples illustrate the limits of governmental ability: governments work best when the rules to be enforced are clear and simple. If you abuse your child in ways that can be easily measured, then you will lose them (emotional abuse, for instance, is a lot harder to prevent). Your children must get these grades at school, and demonstrate this level of attendance, or action will be taken.

And it’s much easier to create rules to avoid very bad situations, than to ensure very good ones. Thus governmental interventions remove the worse case-scenarios, and improve the average in some domains – but can’t be the source of child flourishing or dramatic parenting improvements.

Science, save us!

Finally, why hasn’t science provided us with the knowledge we’d need? If science can study the mating habits of dung beetles in exacting details, and figure out how stars are formed, why haven’t they provided parents with the knowledge they want – why is so much useful parenting advice passed down by word of mouth rather than through textbooks and research papers?

For a start, there are many problems with testing upbringing techniques on humans. Our generations last over 18 years, and it would take even longer to judge the ultimate quality of specific parenting techniques. This is too long for effective experiments on humans. And the target goal changes: each generation has a somewhat different aim for their kids, and the world itself will be different. So even if a proper experiment was carried on for the required length of time, its result would be a list of parenting techniques that… would have worked well three decades before.

Add to that the ethical implications of experimenting on humans, especially over the long term (think of the lack of privacy that would be needed to properly compare the outcomes of different parenting techniques), and it’s no surprise that science has provided many short term answers but few long term ones. So we have many studies on the short term effects of violent video games, for instance, but no consensus at all on the long term effects. And this in a very specific, narrow field of child development.

Furthermore, a lot of parenting techniques are procedural, rather than declarative. Reading a textbook won’t help you, if you’re trying to keep calm as you’re woken by a crying, shit-smelling baby for the fifth time that night – that has to be learnt. It can be learnt through experience, through didactic interactions with previous parents, or maybe through having read the advice and mixing in some practice to get the true understanding. Science has a harder task measuring procedural skill. And if it does find some procedural skills work well, we still have to figure out how to get those skills from an abstract research paper into most parent’s psyche.

Finally, even when the problems and solutions are clear, our technology may not be advanced enough. Some sort of soft, warm robot nanny, that cradled the infant and fed it as needed, would be a boon to most parents, allowing them to focus on more quality interactions with their children. That would be great – but designing such a nanny is a very hard problem in robotics and AI. Sometimes, we know what would help, but we just can’t put it together.

Article dedicated to the imminent Maia Athena Illysia Armstrong, who I aim to bring up as well as we currently know how to.

  • Facebook
  • Twitter
  • Reddit

23 Responses to Why are we not much, much, much better at parenting?

  • Gary Unruh says:

    Here’s a declarative parenting statement coming from the field of child psychotherapy that works every time: Validate a child’s feelings before correcting behavior. Another declarative statement: Validated feelings are a self-confidence cornerstone. Another declarative statement: Our society does not value feeling acknowledgement and sharing. That’s what I know to be true after having the privilege of counseling over 3,000 precious children and their parents.

    • Stuart Armstrong says:

      And the effects of that advice for parenting is… useful but incremental.

      • judith jones says:

        “And the effects of that advice for parenting is… useful but incremental.”

        Useful AND incremental :)

  • Anders Sandberg says:

    Given that we now know that a good rearing environment has pervasive effects on later development and life outcomes (enriched environments, the maternal presence-HPA activity link, prenatal choline improving memory, etc.) , it seems that the importance of improving parenting would rank high on the to-do list of anybody wanting to make the world better. 134 million babies per year times around 80 years of life makes a lot of QALYs – even a 0.1% improvement of life quality due to good parenting means ten million. Non-consequentialists will no doubt regard proper parenting that help produce an autonomous person who can live an excellent life one of the worthiest goals imaginable.

    I think one problem is that just like the economy and schooling, too many people regard themselves as experts since they have experienced the phenomenon. There are deeply seated, loud and often socially sanctioned views that are likely full of cognitive biases and plain wrong ideas. Looking back at the history of child-rearing shows many practices that we today would say are objectively wrong (like swaddling), yet I am confident that parents back then would have defended them strongly. In our own expert-saturated era people act as consumers of expert advice, essentially choosing who to listen to based on not very good reasons. So I suspect the same epistemic flaws that make markets bad at providing good parenting tools are at work here.

    To me, this suggests that the way forward would be to look at ways of making parenting epistemology better:

    * Improving the abilities of parents to be smart consumers is to some extent happening, but clearly it might be worth pushing it more.

    * I like the idea of using life recording and big data to get more real-world input into the process, but Stuart is right about the time lag problem – it will only help for relatively short term, common causal sequences, not slow-moving phenomena you have to wait until late in the child’s life to evaluate, or rare occurrences that provide too little data. There might be some hope for finding mesoscale parenting tips: if your child is in this particular group and you have that situation, do this: one size might not fit all, but could fit many.

    * Maybe we should also investigate how to make useful advice not only detectable as being useful and true, but also figure out ways of learning how to follow such advice.

    * The procedural vs. declarative problem is even harder: knowing useful things like Gary’s advice doesn’t mean one can apply them. Theory and practice are rather far from each other in human life. But any solution to making declarative advice more procedural is going to be far, far more powerful than merely improving parenting, despite the vast importance of the practice. The reason is that such a solution would allow us to multiply human capital more cheaply, improving the world on a vast scale.

    • Tracy W says:

      Actually swaddling is back in fashion. When I had a baby in 2011 the midwives in the hospital (a *big* teaching hospital in London) taught my husband and me how to swaddle said baby.

  • Joao Lourenco says:

    It’s not the case that a product always needs demand to get invented. Many (some say most) products create its own demand. Also, the time issue doesn’t imply infeasibility. There are very long term studies which accompany people from birth to death, but I concede they are rare.
    I agree with most of your reasons, but I don’t think they are strong enough to explain why we have avoided research and good policies in such an obviously important topic (obvious in the sense that existential risks aren’t for example). I believe one of the main reasons is pretty simple: parents don’t want to be better parents. Raising a child takes a lot of important resources such as time and money, it involves many sacrifices. And a parent share more genes with himself than with his child. When the interests of the child and the parent diverge, parents will average this in their calculations on what to do, and often the child will not receive the best possible care. I don’t have the time to list them all here, but there are many, many situations where what’s best for the child is not the best for the parent and vice-versa (if you really want I think Pinker cites a lot of them, I might find for you). For example, breastfeeding on a schedule is correlated with low stressed parents and high stressed babies, while breastfeeding at the baby’s will correlates with high stress levels on the parents and low on the baby. I bet this kind of data would disturb any breastfeeding mother on a schedule. She wouldn’t even want to know that, because then she couldn’t no longer defend that she breastfeed on a schedule so that the baby can ‘learn’ regularity or whatnot. So the real answer is: parents don’t want to be better, they want to be the worst possible kind of parents which would allow for the survival and reproducibly of their child. Evolution is an evil god.

    • Joao Lourenco says:

      One interesting thing is that likely the effects of bad/good parenting are transgenerational (since many of the things parenting affects are transgenerational: cortisol levels, cortisol response, oxytocin levels, probability of divorce, mating style and so on). Hence the effects of bad/good parenting should be very big. This also means that if we could somehow achieve only one generation of good parents, things would tend to stay this way for some time.

    • Stuart Armstrong says:

      So the real answer is: parents don’t want to be better, they want to be the worst possible kind of parents which would allow for the survival and reproducibly of their child. Evolution is an evil god.

      Evolution shapes the wants of parents, but parents don’t want the same thing as evolution. Example: evolution “wants” people to have sex, people want to have the pleasure of sex, but evolution doesn’t “want” people to have the pleasure of sex. So, for instance, if you gave people control over their environment, they would not go in the direction that evolution would “want” them to go (see contraceptive pills).

      Also, there’s strong evidence of cultural changes in how parents relate to children – post Rousseau, post reduced child mortality – on scales much faster than evolution can manage.

      • Joao Lourenco says:

        Yes, I agree with you, I might have push things a little further at the end to make my point. I tried to I describe ‘wants’ evolution has shaped on us to achieve its goals. But I reckon this is not something the parents actively want, probably they just experience this subgoal of evolution as decontextualized human emotions, such as the feeling that the child must be independent or whatnot. (And of course, as a transhumanist, I very much agree that the wants of evolution and of humans diverge, and that we must stick with the latter.)
        The case of contraceptive pill is not the same as parenting. I don’t think some new technology has completely reshaped the way parenting is (as your post argues), so the evolutionary shaped behavior still plays a major role.

  • hmmm says:

    What if it is not really a) possible and b)at all useful to be some sort of super-parent? What would it mean, to maximise the comfort of parenting whilst minimising the harm to the child, or the reverse?

    What if parents delude themselves in thinking that they have all that much influence of their progeny’s outcomes. Sure, you can ruin the life of your child, but a level of good enough is clearly very easily attained. Evolution took care of that a long time ago. This means that the child is robust to the parent’s behaviour, for good or ill.

    Were I to learn that it is possible to be a qualitatively better parent, I would be very much surprised.

    • Will says:

      Anecdotally, some of the most independent and confident children I’ve dealt with had parents who would be considered barely competent parents. While some children with superinvolved parents willing to spend any amount are essentially unable to function on their own.

    • Stuart Armstrong says:

      >What would it mean, to maximise the comfort of parenting whilst minimising the harm to the child, or the reverse?

      Great improvements on either axis would be good!

      >What if parents delude themselves in thinking that they have all that much influence of their progeny’s outcomes.

      The question is not whether current parenting has much of an influence, but whether it could have a lot more influence – and why we haven’t discovered a way to do that yet.

      • fivegreenleafs says:

        “…but wondering if there were ways for parents (broadly construed, to include eg governments) to have great influence, and why we were so terrible at figuring out these methods, if they did exist”

        Firstly, I would urge you to really do read “The Nurture Assumption”, or at least the original article: Harris, J. R. Where is the child’s environment? A group socialization theory of development. Psychological Review, 102, 458-489, (1995).

        If Harris theory do stand up to scientific scrutiny, it does eminently explain the “why”, and point to both the possible opportunities and constraints for the “how”.

        Harris does touch extensively upon these issues as well, both directly and implicitly in her book.

        “…and why we were so terrible at figuring out these methods, if they did exist”

        Were we, are we? To whom do you refer, and on what time scale, historical epoch?

  • Wonks Anonymous says:

    There is a glaring problem with this essay: no mention of Judith Harris’ “The Nurture Assumption”.

    • fivegreenleafs says:

      “There is a glaring problem with this essay: no mention of Judith Harris’ “The Nurture Assumption”.”

      My thoughts exactly.

      If I remember correctly, Pinker in his foreword, stated that he thought “The Nurture Assumption” would turn out to be one of the most important and influential books in the 21th century in psychology, (If my memory does not fail me).

      I would also like to point out some obvious potential reason why we “appear” not to have progressed,

      * Because we have disregarded much cultural “knowledge”, gathered over generations from mothers who did have more the 3 children
      * Because the whole area of child rearing during large parts of the 20th century have been dominated by politicians and political ideas or science informed by diverse ideologies
      * Because childrearing is something you have to “learn”, through experience, the same way you can not “read” or “study” how to become Tiger Woods, or buy a cheap machine at the closest supermarket. In contrast to almost everything else around us these days, it also still takes as much time, dedication, patience it did 100 years ago, if not more…

    • Stuart Armstrong says:

      Thanks for bringing that up! The Nurture Assumption, as I understand, demonstrates that parents don’t have much influence on their children’s outcomes (baring abuse and other excesses).

      I wasn’t looking at the current setup, though, but wondering if there were ways for parents (broadly construed, to include eg governments) to have great influence – and why we were so terrible at figuring out these methods, if they did exist.

  • Kian Mintz-Woo says:

    Our generations last over 18 years, and it would take even longer to judge the ultimate quality of specific parenting techniques. This is too long for effective experiments on humans.

    Although it is true that specific parenting techniques cannot be tested in this manner, longitudinal studies can provide us useful information about what environments tend to be flourishing environments. Of course these are not the blinded/double-blinded tests which we view as the gold-standard. But they are instructive.

    For instance, we can come to tentative conclusions about structures of families such as the children of lesbian parents (in short, that they perform and/or outperform their peers in almost all categories: social, academic, mental health, etc. One striking result was that none of the children reported abuse): http://www.nllfs.org/ This type of social science data does not teach us how to raise children, but it does tell against traditional thoughts that same-sex households are damaging to children.

  • Kian Mintz-Woo says:

    “Our generations last over 18 years, and it would take even longer to judge the ultimate quality of specific parenting techniques. This is too long for effective experiments on humans.”

    Although it is true that specific parenting techniques cannot be tested in this manner, longitudinal studies can provide us useful information about what environments tend to be flourishing environments. Of course these are not the blinded/double-blinded tests which we view as the gold-standard. But they are instructive.

    For instance, we can come to tentative conclusions about structures of families such as the children of lesbian parents (in short, that they perform and/or outperform their peers in almost all categories: social, academic, mental health, etc. One striking result was that none of the children reported abuse): http://www.nllfs.org/ This type of social science data does not teach us how to raise children, but it does tell against traditional thoughts that same-sex households are damaging to children.

  • fivegreenleafs says:

    “Why are we not much, much, much better at parenting”

    First, It could very well be because, one of the absolut most fundamental assumption about child development and child rearing, have been completely wrong.

    As Judith Rich Harris argues for in her “The Nurture Assumptions”, parents behaviour appear to have no lasting effect or impact on long term personal characteristics.

    What the evidence gathered from longitudinal adoption and identical twin studies seems to support, is that for many traits, 50% of the variance can be explained by our unique heritage (i.e. the genes) and the other 50% by enviroment, but NOT, nota bene, the parents and the home environment, but rather the childs peer group.

    Second, Since much of the scientific work within psychology and sociology etc spanning the 20th century did not take genetics into account, many of the results are seriouslly erroneous.

    This might to some degree begin to explain, why, we have not seen any radical “improvements”, (what ever we might mean with improvements), because the prominent ideas have been built on incorrect foundations, and what must today be viewed as, inadequate scientific research.

  • Mark Dionne says:

    Perhaps people could get some good practice for child-rearing by getting a dog and bringing it up properly. (I see a pretty high rate of failure there too.)

  • Tim Conway says:

    Interesting hypothesis that all these exogenous factors might improve parenting. Unfortunately endogenous issues, like rampant narcissism and too many parents more interested in being friends with their kids than parents, trump exogenous.

  • Douglas Brooks says:

    I spent most of yesterday driving through an Amish community in Upstate New York. I was shopping at their farmstands and talking to various craftsmen. I know the community pretty well, having worked with them and spent a fair amount time there. They all have huge families, up to a dozen children, and what I notice is how even toddlers are instinctively drawn to the work of the shop and the fields. I saw many of the children playing, but they did so near where their parents were working. They are all quiet and calm, and also curious of me but seemingly comfortable and amazingly self confident. I had to ask directions of some children who were driving a pair of draft horses and taking a load of manure to the fields. Beyond the amazing amount of responsibility they had been given, the answered my questions and even chatted with me about where I was from.

    Here in Vermont I live in a community full of children I rarely see. Many children are not allowed to even play in their own yards, much less ride a bike down the sidewalk. You can stand in front of the local elementary school and actually see cars pull out of driveways, bringing children just two or three blocks to school. The crossing guard has told me the number of children who actually walk to school has plummeted. I once saw a girl walking to the town pool and her mother was following her in the car…..

    And these local children eventually grow and are allowed some freedom, and I find that they cannot make eye contact, generally speak to me in a monosyllabic fashion, etc. I see fear-mongering on the part of parents who are pathologically risk-averse and an enforced helplessness among the young, who continue to turn to their parents (via cell phones) for daily advice well into college and beyond.

    Among the Amish you meet 24 year old fathers and mothers, running sawmills and large farms, taking on enormous responsibilities and large families. The level of self confidence across all ages is noticeable. To reference Will’s comment above, asking one’s child to lead draft horses, haul manure, plant potatoes, run a farm stand, etc. would be considered child abuse in our world.

    I don’t mean to glorify the Amish; they have their problems to be sure, but I also believe there has only been one documented murder among the Amish in the United States in the last one hundred years. They must be doing something right…..

    • fivegreenleafs says:

      That is a very interesting observation, that illuminates one aspect about much of the historic research done in this area in sociology and psychology, that (more or less), has exclusively been based on western (european) societies.

      Often, only different socioeconomic groups have been compared in studies, (if I have understood the situation correctly), in most cases without no real comparisons between radically different cultures and societes.

      I think this points to a potential very serious defect, (together with the absence of genetics), since antropological data and, even simple anecdotal observations (as you present here), indicate easily observed and obvious apparent differences.

Recent Comments

Authors

Affiliations