Are kids normal?

klio2582/Flickr/CC BY-NC 2.0
klio2582/Flickr/CC BY-NC 2.0
Written by Matthew Woolf

There is a debate that started in 2011 among prominent economists on sites ranging from Cato Unbound to the New York Times Economix Blog. It revolves around a question that may seem out of the scope of economics: are kids normal? When economists use the term “normal,” we are referring to a specific classification of goods. “Normal goods” are those whose demand increases as income increases (cars, food, and most other goods). This is in contrast to “inferior goods” whose demand decreases as income increases (things like public transport and payday lending).

So what type of “goods” are children? Those on “team inferior” – Freakonomics, NYT Economix, Nova Workboard, among others – lean heavily on the well-established negative relationship between income and fertility (Figure 1). The line of thinking basically goes: increasing incomes lowers the demand for children, implying children must be inferior. But is that enough to determine a causal relationship? Those on the “normal” side of the debate – Cato, Cato, Modeled Behavior, among others – firmly say no, citing other key variables like education, IQ, and mother’s age at time of birth.

Figure 1: Fertility rate vs GDP per capita

Short of an honest experiment, a better understanding of the economic theory behind the data is the best way to settle this debate. In the following, we’ll explore how Figure 1 is indisputable, yet a misleading piece of evidence to classify children as inferior. To understand why, we have to consider how the relative costs and benefits of child-rearing differ between developed and developing countries.

What about the costs?

When economists argue that children are inferior goods, they tend to focus exclusively on income changes. They point to the downward sloping curve, whether it be multiple countries or different cohorts within countries, and conclude that children must be inferior goods since poor people have more of them. But assuming away all changes in costs as a country develops makes this reasoning flawed. Costs – both opportunity costs and material costs – also change dramatically throughout the development process.

First let’s consider the biggest opportunity cost to having children: work. Children and work are both incredibly time-consuming so it is a dilemma that parents, especially women, grapple with. This dilemma gets even tougher when work offers a high salary, benefits, and a sense of fulfillment. As an economy develops and jobs offer higher wages and benefits, individuals faced with that dilemma will choose to spend more time on their careers than child-rearing.

Now let’s consider material costs, since the “production” of children is not only a function of time but also material inputs. In the developed world, these costs include school, healthcare, insurance, babysitters, food, shelter, and more. In the developing world, many of the same materials are associated with child-rearing – food and shelter, for instance. But others – school, healthcare, insurance, university – aren’t always. Furthermore, these additional services needed in the developed world are expensive and actually get even pricier as an economy develops.

In Australia, for example, the average cost of raising a child from birth to the age of seventeen is over $297,600. This is a 45% increase in the past ten years while incomes have only increased by about 23%. Compare this to India or China where average income has just about doubled in the same period. While data isn’t available on costs of raising a child in these countries in 2018, it is hard to imagine that they outstrip such rapid income growth. If we take the basic CPI statistics as a rough indication of the cost of children, these numbers have increased much less than incomes.

So costs, opportunity and material, can both lead individuals towards having less children. Not only will development cause individuals to value their time differently, but specific material costs required to have children in developed countries can make them relatively more expensive.

And the benefits?

Just as the relative costs of having children increase as an economy develops, the benefits change as well. It goes without saying that there are a lot of reasons people have children, be it biological, social, religious, or a result of poor access to contraceptives. Moving beyond those reasons and isolating the economic benefits can help us understand the underlying economic motives for having children. Let’s take a look at these benefits and how they differ between developing and developed countries.

In the developing world, there are very tangible financial incentives to having children. Often, they are an insurance policy against old age where pensions and government schemes are nonexistent. The young care for the old as they age without any other financial assistance. Or, children are seen as a way to eventually earn more income for the family. In rural areas an additional child can mean an additional farm laborer.

In rich countries, the tangible economic benefits are virtually nonexistent. Parents may choose to have children for less tangible reasons such as a sense of purpose, or even, as some have suggested, a status symbol. Besides these amorphous reasons, individuals in rich countries reap less personal financial benefits from having children overall.

But suppose you could equalise the income and costs between an individual in a developed country and an individual in a developing country. Despite this leveling, their preferences are still very likely to differ because of the differences in benefits described above. The individual in the developed country will likely have less children, as they can rely on the government for financial support in old age and urbanisation has diminished the benefit of having extra hands to work the farm. Conversely, the individual in the developing country will prefer to have more children to ensure they’re cared for in older age and to provide another source of family income.

And the data?

The theory behind the cost and benefits of child-rearing shows it is possible that children, despite Figure 1, are normal goods. But obviously, theory only gets us so far. What about the peer-reviewed empirical research? Overwhelmingly, the research vindicates these theories. Many papers – Lindo (2008), Lovenheim and Mumford (2013), Black et al (2013) – have come to the same conclusion: children are normal.

It is beyond doubt that fertility decreases as a country develops. Notwithstanding, an increase to individual income will lead not just to more sports cars, organic produce, and designer clothes, but maybe even a new child or two.

image_pdfDownload pdf

Matthew Woolf

Matthew Woolf is a PhD student in economics at the Crawford School, ANU.


  • I am trying to understand the policy implications of categorizing kids as normal or inferior goods. What of children are inferior goods, how does it impact the policy making in developed countries?

  • Hi Matthew,

    That’s a really interesting post thanks.

    Sorry, I’m not an economist and so find the theory you put forwards a little confusing, but aren’t you arguing that *in theory* kids are inferior goods?

    Setting that aside, it seems to me the nub of the debate you link to is whether there’s a positive relationship between income and children *once you take many other factors into account* (Caplan) or whether this is largely irrelevant because its those many other factors that have driven the massive changes we are currently witnessing globally (Woolfers).

    You provide good evidence for a positive relationship once other factors are taken into account with the papers you link to. But it’s worth emphasising — particularly for people working in development policy — that this doesn’t mean other factors don’t matter, or even that other factors aren’t more important in driving the big historical changes. Indeed, the real ramification of kids being normal, once everything else is taken into account, is that other factors must matter much more, otherwise we would have seen fertility skyrocket as countries became more wealthy.

    As an aside, it’s hard to look at Figure 1 and conclude that the drivers of fertility mightn’t be very different in wealthy countries from those in developing countries. And, also that this might be more of a step change rather than something that exists along a continuum.

    Thanks for an interesting post.


    • Thanks Terence! Glad you found the post interesting. And to be clear, while I am an economist and therefore know some of the theory, part of the reason for me writing the post was because I find the debate somewhat perplexing myself and wanted to hear from others. So I approach it humbly – at least as humbly as possible for an economist!

      Theoretically, when considering whether a good is normal or inferior, it’s only the change in budget line, or income, that we are concerned with – prices, costs, benefits, etc, remain constant. So if a good is truly normal (inferior), an increase in income, with everything else constant, would result in more (less) demand for that good.

      The point of my post is to argue that just because an increase in income is associated with a decrease in childrearing (Figure One), we cannot assume that prices, costs, utility, etc, remain constant throughout the development process. So actually, *in theory* children are normal while in reality they appear to not be normal because assuming everything else constant is a flawed assumption. But, if we could fairly assume the constancy of these other variables we would expect to see an increase in fertility.

      I think I grant your point about other factors being more important for such historical change since we don’t see fertility skyrocketing in rich countries. Maybe the best way to approach it is not to equate development – where the constancy of other variables cannot be safely assumed – with exogenous income increases – where they can be safely assumed. Obviously, the vast changes experienced by a country going from poor to rich, as shown in the fertility data in countries that have done so, is vastly different than a country or region experiencing an exogenous one-off income shock, as shown with a coal boom in Appalachia USA in Black et al (2013) where fertility did increase. In the first case, there is obviously a lot more happening and while children may still be theoretically normal, we can expect to see a negative income-fertility relationship because there is just so much changing. But when we look at a mining boom in Appalachia or perhaps an iron ore boom in WA, we would expect to see an increase in fertility. Still, economists seem to have mixed views about all this which is why I have been exploring it.

      Finally, I want to argue why I think the question of whether children are normal or inferior may be less intellectual pontificating and actually important (a partial response to Ashlee’s last post). Many macroeconomic models focus on k=K/N, the capital per worker in an economy as the engine of economic growth. As an economy develops – or, as K and N both increase – they will be competing with each other to determine the way k changes over time. If fertility is increasing at a low (high) rate, k grows faster (slower). I think most people would agree that fertility is an important indicator in an economy yet we still don’t seem to fully understand how it changes with changes in income.

      Of course, this sort of modeling is usually done on aggregate so the individual cases Ashlee mentioned in her previous posts are too often replaced in favor of a more simplistic “representative agent” model. I cannot defend every flaw of economic theory and I do not deny the importance of these other factors. Obviously this is imperfect and I’ve actually been trying to crack this with some sort of heterogenous agent model that considers two or three cases, but it’s really hard – maybe even impossible.

      Let me know if that clarifies the post a bit. If not, I’ll do my best to try again.

  • While I understand that economists like to pontificate on this stuff and there is some value to it, sometimes it kills me and makes me think there should be more women in economics! While yes there are huge work-related ‘costs’ in having children, more heavily borne by women, there are also quality of life costs.

    There’s the costs of pushing a bowling ball out of your body which are solely borne by women. There is a risk of permanent or temporary disability from the act of childbirth. Also, I am told that it hurts a lot to give birth and it is uncomfortable to be heavily pregnant! For women who breastfeed, this is another time cost borne solely by the woman — a form of economic production really that is completely ignored.

    I also feel that economists also conflate demand and supply of children — the fertility rate measures births, ie supply, not actual ‘demand’ for children, marginalising people who perhaps have experienced infertility/pregnancy loss, or who feel they can’t afford to have children/as many children as they would like. There’s also other groups for which the economic costs of falling pregnant are substantially higher — it’s very expensive for lesbian couples or single women to go through the IVF process, for example, or for couples experiencing infertility. These opportunities are only available when income increases.

    There’s also the issue of the undervaluing of women’s domestic labour, and the ‘second shift’ issue for working women. An increase in income may mean the ability to outsource some of this labour, which makes balancing work/home easier, for example. But in a developing country, where women’s labour is informal, this ‘cost’ often seems overlooked.

    One clear trend is that when given a choice, women choose to have fewer children. Not just because of job satisfaction/career, but because of bodily autonomy and a whole host of other reasons — in the US, the crushing burden of educational debt is being cited as one reason that Millennial women are putting off childbirth, as one example. It’s complicated! In developing countries women often don’t have access to the same choices due to gender inequality.

    So I don’t think you can de-couple a discussion about income and reproduction from women’s rights, and their place in an economy. I understand there are some clear differences between the developing/developed country fertility rates etc, and there is a fascination in considering why and what the economic trend is etc, but kids sometimes seem the most irrational consumer decision of all! Especially for women who bear the biggest costs by far. (They are so cute though…)

    • Also, in terms of an economic good, normally you achieve efficiencies in production through ‘specialisation’ — there’s very few avenues for this for women in reproduction. Outsourcing pregnancy to a surrogate (huge amount of ethical issues) would be the main one, again usually only accessible to those with high income (Kim Kardashian has done this, for example, due to suffering placenta accretia — most women with this condition would either have to stop having children or gamble on extremely high risk pregnancies). ‘Economies of scale’ also has its limits in this form of production… Octomum has incurred some pretty significant additional costs, for example. I don’t think you can really rationalise this one!

      • Thanks for the comments, Ashlee. I agree with the following:

        1. There should be more women in economics: Enough said.

        2. Issues beyond economics – women’s rights, health, infertility, and others – play an important role in fertility: I tried to acknowledge this a bit in the ninth paragraph: “…biological, social, religious, or a result of poor access to contraceptives.”

        I disagree, however, with the following:

        1. Looking at economic factors alone isn’t useful: You don’t say this explicitly so apologies if I am mischaracterizing, but anthropological research – – has shown that out of sixty-four factors considered relating to family size, the economic ones were most important. This is somewhat vindicated by survey data that says families would have more children if they could afford it. In light of this, it seems economic forces are, if anything, underrated in understanding fertility.

        2. Economists conflate demand with supply: Economists also model fertility using childrearing production functions. It doesn’t seem like one approach is necessarily favored over the other – both approaches are well-represented and are useful. But you actually kind of make the case for considering the strengths of using a demand approach when you point out how specialization and economies of scale don’t really make sense with fertility.

        On your last point, I am definitely not trying to rationalize the decisions of Octomom. That doesn’t mean that economics still can’t rationalize – with some degree of accuracy – fertility choice for most people.

        That also doesn’t mean that there isn’t a place for discussion about the inequalities and differences existing within fertility choice. As with any aggregate analysis, there is more going on underneath that can be unpacked. But that’s just not what this specific post is about. I’m curious though, if you were to only consider economic factors, would you think children are normal or inferior goods? Interested to hear your thoughts.

        PS – babies are sometimes cute…

        • Thanks for your reply Matthew. I certainly see the utility of economic analysis on family size choice etc and other factors, but as a non-economist I just get frustrated by what are counted as ‘costs’ — the bodily toll on women etc often being ignored, and their unpaid domestic labour and reproductive labour. I also question why there is a need to categorise a child as a normal/inferior good for analysis purposes when it very much dependent on context and circumstance!

          From a purely economic standpoint, as a single woman in my mid 30s in a developed country, a child for me would be a normal good, because I would have to go through very expensive IVF so there would be an outlay of capital, and also to afford to pay childcare to continue to earn an income (high recurrent costs). (Yes, I could just randomly hook up with someone, hope it worked, and skip the outlay, maybe quit my job and live on low income on government payments, but we are talking about consumer preference, so…. )

          But I have full reproductive autonomy, something many women globally don’t have. This to me is key.

          Even in a developed country, say someone gets pregnant by accident, a child could be an inferior product. For example, if they can’t afford to access a termination, whereas someone on a higher income would be able to. Or if they couldn’t afford/access contraceptives. Then the child is a direct result of the constraints of living on a low income. (And again, this example is about the supply of children, not necessarily the demand!)

          So for me it goes either way, and a huge part of it depends not just on income but on the choices available to women and policy and social settings around reproductive rights and gender equality more broadly.

Leave a Comment