Yes, the decision to try (or not) to have a child can be made rationally

Philosopher L. A. Paul and sociologist Kieran Healy write:

Choosing to have a child involves a leap of faith, not a carefully calibrated rational choice. When surprising results surface about the dissatisfaction many parents experience, telling yourself that you knew it wouldn’t be that way for you is simply a rationalization. The same is true if you tell yourself you know you’re happier not being a parent. The standard story of parenthood says it’s a deeply fulfilling event that is like nothing else you’ve ever experienced, and that you should carefully weigh what it will be like before choosing to do it. But in reality you can’t have it both ways.

I disagree that you can’t have it both ways, for three reasons:

1. Many potential parents do have an idea of what it will be like to be a parent, having participated in child care as an older sibling, aunt, or uncle.

2. The decision of whether to have a child occurs many times: the decision of whether to have a second child, a third child, etc. All these decisions after the first child are directly informed.

3. Finally, and with most direct relevance to Paul and Healy, even though you can’t know how it will feel after you have the baby, you can generalize from others’ experiences. People are similar to each other in many ways, and you can learn a lot about future outcomes by observing older people (or by reading research such as that popularized by Kahneman, regarding predicted vs. actual future happiness). Thus, I think it’s perfectly rational to aim to have (or not have) a child, with the decision a more-or-less rational calculation based on extrapolation from the experiences of older people, similar to oneself, who’ve faced the same decision earlier in their lives.

All three of the above points are pretty obvious so I assume that Paul and Healy have thought about them, but I didn’t see these issues mentioned in their blog post or the linked article. So I’m raising them here. And here is as good a place as any to emphasize that I’m not trying to “debate” or “debunk” Paul and Healy but rather to integrate their ideas with other things I know about.

Humans: Rational animals or irrational computers?

I’m sensitive to this particular issue because I see Paul and Healy’s article as taking what might be considered a generally “humanistic” position that there are limits to the value of rational calculation. But to me, as a statistician, there’s nothing particularly humanistic about disparaging rationality. Rational thinking is, to me, a key part of what makes us human.

To put it another way, it used to be that humans were defined as the rational animal. Nowadays it almost seems the opposite, that we are defined as irrational computers. When the baseline is dogs, cats, monkeys, etc., we indeed look rational (even after accounting for the fact that animals behave rationally in many situations). We don’t (always) act on instinct, we make plans, etc. When we are compared to computers, the story is very different, of course. But we’re not computers, we’re animals [I apologize to Rick Santorum and any other creationist readers of this blog], so I think our (imperfect) rational thought and behavior is nothing to shy away from. Indeed, even Paul and Healy, when denying that we can be rational in setting childbirth aims, do acknowledge that people are following thought process that have the form of rational reasoning, they’re just claiming that the inputs to this reasoning are empty of real content.

The aim (not decision) to have a child

I write “aim” to have a child rather than “decide,” because many people who want children, can’t have them, and many other people who don’t want children end up with them anyway. So you can’t quite “choose to have a child.” In statistics jargon, this is all an intent-to-treat analysis.

41 thoughts on “Yes, the decision to try (or not) to have a child can be made rationally

  1. 1. I like the sensitive use of “aim”!

    2. I certainly recall a number of long discussions with my spouse about this topic. I am currently watching/kibitzing with my daughter and son-in-law on this topic as well (although now more about timing than yes/no).

    3. Parenthood works out differently in practice than you think it would, often. Jorge Cham has a nice “statistical graphic” on this http://www.phdcomics.com/comics/archive.php?comicid=1563 But that’s not necessarily much different than macroeconomic decisions not quite working out the way we thought, or the well-known principle of unintended consequences. In fact, if you want ‘leaps of faith’ macroeconomic fixes clearly involve more faith and more uncertainty about results.

  2. Horses are rational: They have complete and transitive preferences over sugar > carrots > hay. I believe parakeets are rational too. Sometimes rationality is “too weak” an assumption.

    Perhaps more appropriately we might say humans are the only animals who think they alone are rational.

    • But rational preferences are not neceassary and not nearly sufficient for rationality. Rationality is a question of the choices you implement to fulfill those preferences. Do horses eschew carrots now for (enough) sugar in 20 minutes? One can also be rational without complete and transitive preferences. I do not know the marginal rate of substitution for me between being better at chess and owning a few square miles of Jupiter, but fortunately I’m at a corner solution for both. And intransitive preferences don’t cause any particular problems so long as you’re rational enough to spot the Dutch books coming at you.

      • Jonathan:

        Anecdotal evidence suggests that when presented with any pairing from carrots, sugar, and hay, horses make the necessary choices to fulfill their preferences as stated above. Try it next time you go to a petting zoo.

        PS you are spoiling the story. Yes, acyclic preferences will do; and we are interested in rational choice; and we should be talking about revealed preferences; and we can write a book on what it is to be rational. But my point is not to say that horses, parakeets, rats and opossums are rational tout court, but simply to point out that rationality is sometimes blown way out of proportion.

    • My pet rabbits, who live in the back yard, seem to come with built in random decision generators. After eight years, I still can’t predict when I come out to feed them whether they will dash off, ignore me, or come running to me. I suspect this is helpful in not getting eaten by predators. They aren’t very smart but random decisionmaking makes it hard for predators to outsmart them if they don’t behave rationally.

  3. Many times having babies is the default transition in the life course. … school,job, marriage, babies… somehow an automatic response to institutional arrangements.

  4. I didn’t understand that article at all. It also seemed to ignore that the number of people deciding to have children (and the number of children they are deciding to have) is changing rapidly; that is, people are making the decision based on some basis that is connected sufficiently to the external world so as to be changing for hundreds of millions of women simultaneously.

    Incidentally, is there good empirical research about the extent to which men are actually making decisions about children at all? I think demographers usually talk about this as if it is solely women’s empowerment, choice, and available alternatives which drive changes in fertility, and I was curious if we can ever observe men’s preferences changing as well as a shifting equilibrium between men and women’s revealed preferences.

  5. I would like to add that the decision to have or care for children does not have to be just about how you yourself will enjoy the process. Instead, it can be a moral choice. Before my daughter was born I knew I would not much care for the work of raising a child and did not need the love or support in my old age. I also felt exceptionally fulfilled before she was born, and so expected little improvement in that area. The reason I decided on a child is because I believe more human life is better and that she would appreciate my choice. For this reason, and this reason alone, I plan on about 3-4 children.

  6. Agree. Like Frank Knight before them, Healy and Paul set up a false dichotomy between settings characterized by mere risk (the payoffs and the probabilities assigned to these payoffs are known with certainty) and those characterized by uncertainty (one has no way of knowing what is going to happen). In fact, decision situations can be located anywhere on a *continuum* of uncertainty. The most charitable reading of the Healy-Paul argument is that they are trying to say that having children (or not) is a decision that’s pretty close to the “perfect uncertainty” end of that continuum, and that they are correct in saying so as far as the first child is concerned.

  7. Thanks for your comments, Andrew. The blog post is really just a summary of my paper on this. I address objections (1) and (3) in the paper, although perhaps not to your satisfaction. And I think your point (2) raises interesting issues that I don’t have firm views about.

    Just a bit more regarding your comment (1). My main concern is with one’s ability to assign value to the experience of what it is like to have one’s own child. In the paper, I argue that the process of creating the child through pregnancy and birth gives the experience of having one’s own child unique properties, and that babysitting and such is different in kind from having a child of one’s own. (Thus the various memes about “other people’s children,” including those about how one can dislike other people’s children while loving one’s own, about how adopting a child “isn’t the same” as having one, etc.) Experience with other peoples’ children might teach you about what it is like to hold a baby, to change diapers or hold a bottle, but not what it is like to create, carry, give birth to and raise your child. In fact, there are purely biological causes that are sufficient for the uniqueness of the experience: the hormonal reactions and other biological responses that stem from physically growing, carrying and giving birth to a child. And while this argument obviously targets biological mothers, I think similar arguments apply to biological fathers–they even have biological responses. (What I say may apply to adoptive parents as well, but I don’t have enough empirical facts to say much about this one way or the other.)

    There is more to say about (3) but this reply is getting too long already.

      • OK. With respect to point (3), there’s a lot to be said. Maybe the best place to start is with the empirical research. The psychological research almost uniformly shows that those who never have children report higher levels of happiness than those who do have children, at all stages of life. If we accept this research at face value, and extrapolate along the lines you suggested above, it suggests that nobody, if they are choosing rationally, should choose to have a child. This result is obviously catastrophic.

        So take my point to be something like this: if we follow the ordinary approach to choosing to have a child, we can’t choose rationally, due to the special epistemic situation we are in prior to having the experience. If we revise our decision procedure to make the choose rationally, we get bizarre, unacceptable results.

        My view is that decision theory is great most of the time. But for certain “life-changing” decisions, it may not be appropriate. In particular, when making these sorts of decisions, we may need to recognize that we have certain epistemic limits and thus, in effect, our choice amounts to a leap of faith.

        • LA:

          Yes, I’m aware of this research. Nonetheless, I don’t see this as refuting the decision-making approach I suggested, of looking at older people and seeing what their lives are like. It’s not just about moment-to-moment happiness, it’s also about life satisfaction. A young person can look at my life and see that my kids take a lot of work but that they obviously give me a lot of satisfaction.

        • I believe the psychological research shows that even “overall life satisfaction” is lower for parents. It might be the case that the research is flawed. But I’m hesitant to assume that the research must be flawed simply because the results conflict with my gut intuitions.

        • My understanding of the happiness research is that self-reported happiness goes down with children but self-reported life satisfaction goes up, just as you say. I think L A Paul is wrong in his comment to say otherwise.

        • > The psychological research almost uniformly shows that those who never have children report higher levels of happiness than those who do have children, at all stages of life

          That’s so wildly implausible I have to guess you are summarizing it ambiguously. Perhaps you meant to say that _average_ happiness given a broad conditioning set (life stage, children or not) satisfies this pattern. Is this what you mean? Because if so it absolutely does not follow (not even even close) that:

          >nobody, if they are choosing rationally, should choose to have a child.

          People always have a lot more information to take them into a narrower conditioning set.

        • Some of you may find it wildly implausible, but that’s what the research suggests. The whole point of this discussion is that we need to look beyond simple intuitions, gut reactions and naive introspection when thinking about how we reason about major life decisions.

          For a brief summary of the psychology you might look at this post.

          http://www.psychologytoday.com/blog/explananda/201303/is-it-rational-have-child-can-psychology-tell-us

          For some academic papers is psychology and sociology: McClanahan S. and Adams J. (1989) “The Effects of Children on Adults’ Psychological Wellbeing.” Social Forces 68: 124-46. Nomaguchi K. and Milkie M. (2003). “Costs and Rewards of Children: The Effects of Becoming a Parent on Adults’ Lives,” Journal of Marriage and the Family 65 (2003): 356– 374. Simon, R. (2008). “Life’s Greatest Joy?: The Negative Emotional Effects of Children on Adults.” Contexts 7: 40-45.

        • (Replying to L.A. Paul … comment nesting doesn’t work that deeply it seems)

          > Some of you may find it wildly implausible, but that’s what the research suggests

          I read as much as I could find without paying for the specific papers you cite. I read a bunch of papers that cited these. What I did find is consistent with your language, but is a uniformly outrageous abuse of language and complete incomprehension about statistics. These cite _population_ level results about direction: conditioning on a variety of general explanatory factors, some average measure (of happiness) seems statistically higher in one case than
          the other. And yet these papers seem to quickly (as do you) jump to concluding this says something about individuals and their rational choices.

          If I gave 100% statistical evidence, incontrovertible, that people born in (say) August (i.e. just before northern hemisphere school starts) performed “worse” _on average_ than people born in other months (and let us say the average SAT difference was 0.000001%, yet real – i.e. statistically discernible)) would you then conclude that no August-born person, choosing “rationally”, should choose an academic career – since how could they expect compete with all the smarter people out there, indeed anyone in born in any other month is “proven” to be more intelligent. What you are suggesting, and the open psychological papers on parental happiness I could get to, seems tantamount to this error.
          If you mean something else, can you please explain.

          I was being overly tactful in my first note. The claim:

          > The psychological research almost uniformly shows that those who never have children report higher levels of
          happiness than those who do have children, at all stages of life

          isn’t implausible. As written, it’s just false. Obviously so. One parent happier than an one, otherwise demographically non-parent, disproves this. It utterly and surely fails unless you include a (rather broad) averaging process in there, and in which case one is entitled to ask about the magnitude. (Is it 0.000000000000000000001% higher? is
          it 50.0000001% of non-parents in the class who are happier? Or what?)

          If you rewrite it in a way that might be arguably true, then people might be in a position
          to discuss whether your “rational choice” argument follows even with a tiny level of plausibilty.

          In my attempt to follow the literature you point to (and I cannot find a non-gated version of these papers, but many others in their vein that cite them) I am somewhat astonished to see a field that doesn’t seem to grasp the difference between population-level statements about the sign of a difference vs recommendations made for a single case. Amazing.

        • It’s worth emphasizing that the claim in the paper is that the E[U] approach is irrational while the claim that “[i]f we accept this research at face value . . . [then] nobody, if they are choosing rationally, should choose to have a child” leaves three possibilities:

          (1) your claim, that E[U] is irrational (using proof by contradiction)
          (2) your approximate E[U] is wrong, as Andrew notes below
          (3) people make irrational choices about having kids

          (3) is true for many people and (2) is probably also true so it’s a leap to conclude that this is evidence for (1).

        • Steve, you’ve conflated the decision procedure I describe as “ordinary” with the empirically based decision procedure I was discussing with Andrew above. Our discussion above involved using a different decision procedure to make the choice, which I objected to because the result was catastrophic.

  8. Re objection 3: I think one can rephrase the Healy/Paul argument as follows. No one has both had a child and not had a child. People when they have children are psychologically changed in such profound ways that a childless person doesn’t know what it’s _like_ to be them any more than what it’s like to be a bat. Childless people know that people “like them” — in the sense of being from the same class and neighborhood — have had children, but these people are _now_ as unlike them as bats, so whether they’re happy is not useful evidence. (It is like one of those thought experiments where you have to decide whether to undergo a procedure that ruins your brain but makes you “happy.” Obviously the negative answer is defensible in this case; it follows that the yes answer to the child Q. cannot be justified on the grounds of reported happiness.)

    I’m genuinely very confused by your “rational animals” point. Definitely one can ratiocinate about anything on the basis of false premises and dodgy logic and it doesn’t look much like gorillas do, but that’s not what anyone ever means when they accuse people of making irrational choices. (Unless “irrational” is treated as synonymous with “impulsive”/”automatic” but taking this really seriously probably leads to the conclusion that trained quantitative people do basic algebra irrationally.)

    • Sarang:

      I think that you as a (hypothetical) 20-year-old can generalize based on older people whose 20-year-old selves are good matches to you. There’s still a cohort effect but I don’t see anyone worrying about that, so I’ll set that aside.

      As for the rationality, I’m thinking of it more along the lines of intellectual history—the idea that, from different perspectives, “rationality” can be viewed as a plus or a minus. I agree that the term has different meanings in different contexts.

      • Hi Andrew, The happiness effect of having children seems dependent on many individual-specific details that make it difficult to reason based on the experiences of (seemingly) similar people who have already had kids. That is, we may need to condition on quite a lot of individual-specific characteristics to get a useful prediction of our own future happiness. Furthermore, it’s not easy to judge the happiness of other people, especially in a context where there are strong social norms and expectations at play (which influence how people present their happiness or otherwise).

        • Humans have an innate proclivity to rating their own experiences as somehow special.

          Objectively, I doubt your joy in childbearing is any significantly different from others. We just like to think that it is.

    • @Sarang:

      A gunshot wound can change people in profound ways too. Again, no one has been shot and not been shot all at once.

      Yet, extrapolating from the painful experiences of millions of others I know confidently enough that a gunshot won’t be a pleasant experience.

  9. For the past eight years my mantra has been “One of the best reasons for having children is having grandchildren” — you can guess the age of my oldest grandchild. Which brings me to the point that a 20- or 30- (or 40-) year-old can’t really comprehend what it’s like to be in one’s 70s or 80s, with career mostly behind him/her, on the downslope of life, with perhaps less than perfect health, and without the vicarious enjoyment and heartaches that come from watching your grandkids (and in my case, stepgrandkids as well) meet their challenges.

    • “One of the best reasons for having children is having grandchildren”

      Right. Being a parent is a lot of work. Being a grandparent is a lot of fun.

      I haven’t seen many models of what factors influence the number of grandchildren, although it’s a fascinating question for numerous reasons, including Darwinian fitness. One source of data is New York Times obituaries of prominent people, which always mention the number of grandchildren at the end. I usually root for high achievers to have left a lot of grandchildren behind so the rest of us can continue to benefit from copies of their genes.

  10. Andrew, you argue that rational thinking is “part of what makes us human” but I have also seen you criticize utility theory as “folk psychology”. I’m agnostic in the debate, but I have trouble reconciling your two positions. In economics, utility functions simply encode rational preferences. In fact, as long as preferences are “continuous” – a property you can probably infer the meaning of – rational preferences are necessary and sufficient for a utility function representation.

    Is your problem with the type of utility functions that are used? Or maybe it is with those that interpret utility in the sense of happiness/pleasure? Or something else I have not thought of?

    • Brad:

      I think rationality is part of how we think and behave but only part. One of my criticisms of the way utility theory is implemented is that there often seems to be an attempt to apply it to all of human decision making.

  11. Do creationists not believe that humans are animals? They may not think we’re apes, but we’re certainly not vegetables or minerals.

    • My impression is that creationists are often fairly level-headed about things that more liberal ideologues get fundamentalist over. For example, creationists reject “macroevolution” but tend to be more tolerant of the notion of “microevolution” within species. In contrast, Stephen Jay Gould advised his readers: “Say it five times before breakfast tomorrow; more important, understand it as the center of a network of implication: “Human equality is a contingent fact of history.””

      • Steve:

        Interesting point. Creationism is (at least among its more prominent adherents) associated with political conservatism, and conservatives are generally more interested than liberals in talking about inherent differences between groups of people. In contrast, liberals are more interested in talking how different groups differ in their material resources. For example, conservatives tend to be happy to talk about differences in test scores comparing blacks and whites, whereas liberals don’t like to talk about it, whereas liberals talk a lot about income and wealth inequality, a topic that conservatives tend to minimize.

        Or, to put it another way, whatever his views (actual and stated) on inherent differences between groups of people, Gould would not have denied human inequality of circumstance.

        • Yes, but Gould was supposed to be an expert on evolution, so his demonizing of actual experts whom he hated for ideological reasons had a malign effect on popular understanding of science.

        • Steve:

          I agree. Gould spreading misinformation is bad. I also think it’s bad when major political figures endorse creationism. But I’m not saying this because of some sense of political balance. A is bad and B is bad. The existence of B doesn’t excuse A in any way (or vice-versa). All that said, I also find it interesting to think about the connections between people’s scientific beliefs and their political attitudes.

  12. Pingback: What is it like to be a vampire and/or parent? | Quomodocumque

Comments are closed.