Are children an unwanted luxury?

The latest accessory

An article last week by Mike Konczal went semi-viral by describing a third child as a luxury good.¹[1] Using data from the US Survey of Consumer Finances, Mr. Konczal showed that the top 1% of households by income were more likely to have three or more children than the less affluent. Some were perplexed by this, including perhaps initially Mr. Konczal, though to his credit he explicitly noted that the income-fertility relationship is complex. Many others seized upon his charts as evidence of a fashionable view that the decline in fertility is driven by rising costs of living, particularly young people being left behind in the property market.

So many issues…

There are so many issues to tackle here that I’ll limit myself to four. First, I’ll explain why Mr. Konczal’s finding is not the least bit surprising and is, in fact, the norm in human history. Second, I’ll dispell three misunderstandings that drive at least some of the confusion and misdiagnoses applied to his result. Third, I’m going to debunk the fashionable cost-of-living argument. Finally, I’ll explain why its popularity reflects cognitive dissonance over what I view as the three most likely causes of declining fertility.

Perplexing perplexity

Like surely any anthropologist, biologist or non-Westerner, I am perplexed that anyone is puzzled by one of the most obvious and clichéd truths of human history and evolutionary biology: the rich can (and do) afford more kids. Just as alpha male chimpanzees monopolize breeding rights within their troops, the rich and powerful have done the same in human societies since the dawn of civilization. Incan emperors reportedly received 1,500 women per province, Chinese emperors had harems of 10,000 and Indian rulers were reported to have as many as 16,000 in their harems.²[2] Genetic evidence suggests that Genghis Khan is the paternal ancestor of 0.5% of the world’s male population,³[3] while genealogical studies reveal that wealthy men fathered twice as many children as poorer men over much of English history.⁴[4] That truism has continued into modern times to such an extent the the authors of one study sheepishly stated it thus, “This finding is not particularly earth-shattering: it is no surprise that people assess their material wealth as part of their decision to have (more) children.”⁵[5]

Or, WEIRDly confusing?

What is behind the confusion? I suspect it has a lot to do with what Joseph Henrich called “WEIRDness” (Western, Educated, Industrialized, Rich, and Democratic).⁶[6] Many Western elites live in echo chambers that ring painfully with cognitive dissonance when struck by reality, as I discuss below. (Prepare yourself: it will be uncomfortable for some.) But those echo chambers are reinforced by a few well defined misunderstandings.

Nations aren’t people

The primary source of confusion, acknowledged by Mr. Konczal, is that many wrongly infer within-country income-fertility relationships from cross-country patterns. Rich countries have lower fertility rates than poor countries due to the “First Demographic Transition.”⁷[7] In the first transition — I’ll get to the second below⁸[8] — economic development improves health and living conditions, lowering child mortality rates, which allows parents to achieve the same desired number of surviving children with fewer births. Further, when parents are more convinced their children will survive, they tend to invest more in their education, reducing their desired number of children due to the increase in time and resources required. As a result, national fertility rates fall as economic development rises. Yet, within countries at every stage — developing,⁹[9] industrialized¹⁰[10] or post-industrial¹¹[11] — richer families tend to have more children.

Gender confusion

An ancillary source of confusion is worth touching upon. Fertility rates are measured in children per woman for a reason. The most prolific mother in history gave birth to 69 children in 27 pregnancies(!), while the most prolific father in history (reportedly) sired 1,171 children.¹²[12] While Mrs. Vassilyev’s 69 children (amply) boosted 18th-century Russian fertility, Sultan Moulay Ismail’s monopolization of his hareem likely reduced the fertility 17th century Morocco. Because women, unlike men, are limited to one pregnancy at a time, they determine a society’s fertility rate. But, as Lyman Stone notes, the data Mr. Konczal uses almost surely double-count at least some US women.¹³[13] Mr. Konczal looks at total children per household (not on a tax-claimant basis), including adopted and step children. Under that accounting, my own blended family — which you can assist into the top 1% with a subscription to Thematic Markets! — is blessed by five children, but in fertility terms it’s only 2.5.

The cost-of- living canard

But the virality of Mr. Konczal’s post was driven by a misconception that is a subtle twist on the nation-for-individual error noted previously. Many took Mr. Konczal’s rediscovery of the wheel as evidence supporting the increasingly common misperception that rising costs of living are driving the fall in fertility rates. But the affordability thesis suffers from two glaring fallacies: a failure to distinguish nominal from real purchasing power and its treatment of childbearing as a primarily economic rather than social choice.

Shocker: rich neighborhoods are more expensive

Let’s start with the first. Somehow people have lost sight of the obvious point that local prices are a function of local income. Without higher incomes to support demand, prices would be lower. This is why high-income countries (and cities) have higher nominal costs. But living standards are determined by real (price-adjusted) income, which is generally higher in richer areas. This should be intuitively obvious: why do people flock to expensive cities for jobs? Because they pay higher real wages. Yes, those higher wages drive housing prices and private-school fees higher, compressing much of the real-income advantage. But that returns us to the truism with which we started: wealthier people are better positioned to have the families they want.

Economic choice?

Further, fertility rates aren’t clearly inversely related to the cost of children. Figures 1 and 2 both show the cost of childcare for two children — after state subsidies and tax breaks — as a share of disposable income, Figure 1 for couples and Figure 2 for single parents. Contrary to the cost-of-living hypothesis, fertility rises weakly with at least one direct measure of the cost of childrearing. Admittedly, the causality almost certainly runs the opposite direction: the lowest fertility nations spend the most to subsidize childbearing…but fail miserably because falling fertility is not primarily due economic choice.

Powdered milk, squirrel stew & horse feed

That cost is not the driver of fertility should also be obvious to anyone who reflects on previous generations. I was one of five kids. We shared rooms, wore hand-me-downs, and even drank powdered milk because fresh was too expensive. My mother and her three siblings shared a single bedroom and ate squirrel that my grandfather hunted in the woods of West Virginia. As a small child, her squirrel-hunting father ate rutabagas (horse feed) that he stole from fields outside Hamburg after sneaking out of the (then) walled city after curfew; the one-room tenement he shared with his brother, sister and parents was heated with coke spewed from steam engines he scavenged from along railroad tracks. Despite having real incomes much lower than most Americans enjoy today, my parents, grandparents, and great grandparents chose to have more children. Yours likely did, too.

Hell’s Kitchen versus Suburban Utopia

“But Marvin, they didn’t have to pay for private schools, universities, and today’s housing costs!” I hear you protest. I retort, “Those are choices.” Consider that Manhattan’s population is 38% lower now than a century ago when the Lower East Side squeezed as many 665 people into each acre because most families in the neighborhood then had five or more children.¹⁴[14] But we don’t have to travel back in time; we can also see choice’s impact on fertility in its differentials across urban, suburban and rural communities in advanced economies. Suburban and rural communities have 10-15% higher fertility rates than cities in the US,¹⁵[15] the same average as for Europe, though for some European countries the difference is as great as 40%.¹⁶[16] But intriguingly, a number of studies find that fertility rates are higher in suburbia than either cities or rural areas,¹⁷[17] and that remains true even after controlling for income, education and the choice to migrate from cities to suburbs with lower housing costs and quality public schools.¹⁸[18] A critical factor driving these results is that urban residents choose to have their children later in life, reducing their overall fertility rate.¹⁹[19]

Denialism

My perception is that some of the people leaning on the “cost-of-living” thesis are reluctant to admit that they’re choosing to “live their best lives” over having children. I’m not judging that choice, just making clear that it is a choice. Enjoying the city life in an Instagrammable apartment (with no bunkbeds!) and taking five exotic-locale vacations per year sounds a lot more fun than Olive Garden dinners out, mowing the lawn of your cookie-cutter suburban house, and packing the kids in the station wagon for a road trip to Walley World each summer. Denying the role of choice also evades facing the potential reverse causality in Mr. Konczal’s post: the families with the most kids may have better prepared for their financial future to have kids. Evidence from two studies of the effects of home prices on fertility suggest that may be the case. Even within the same neighborhoods, a rise in house prices boosts the fertility of homeowners while depressing renters’.²⁰[20] No doubt the findings may reflect differences in inherited wealth for some, but it also reflects some families’ prioritization of savings over “best lives” that boosts both wealth and fertility.

The Second Transition and smartphones

There’s evidence for the role of choice in fertility declines in the literature on the Second Demographic Transition.²¹[21] The collapse in global fertility rates across countries in the last decade cannot be readily explained by improvements in healthcare or greater investment in children’s education. Instead, it appears to reflect what the OECD describes as “Changes in attitudes towards parenthood.”²²[22] As popularized earlier this year by John Burn-Murdoch at the Financial Times,²³[23] these changes in attitude towards parenting appear to have been accelerated by broadband and mobile telephony,²⁴[24] particularly among lower-education women.²⁵[25] Yet, evidence from Lyman Stone, among others, shows that the effects are not uniform: in some populations mobile telephony raises fertility while it reduces it among others.²⁶[26] The causal mechanism appears to be that women are better informed about how to achieve their desired number of children.

Cognitive dissonance

The Second Transition debate, especially over how smart phones affect fertility, exposes a strong cognitive dissonance among Western elites. What changes in attitudes towards childrearing is mobile telephony propagating? And what decisions is it clarifying for prospective mothers? Three likely factors that are strongly correlated with fertility often get short shrift in popular and policy discussions because they are totems of Western modernity: secularism, divorce and state pensions. Figures 1 and 2 subtly testify to the avoidance: 50% more OECD countries collect data on single parents’ childcare costs (Figure 2) as do for couples’ costs (Figure 1); and the extreme fertility outlier in both charts is the OECD’s only member state defined by its religion.

Secularism

Among countries in the Second Transition, religiosity is among the most strongly correlated with higher fertility, while secularism correlates most strongly with lower fertility. For example, in the US, Mormon families average 3.4 children, evangelical Protestants and Catholics 2.3, atheists 1.6 and agnostics 1.3.²⁷[27] These religious differences persist after controlling for income, education and urbanity, and are similar to those found around the world.²⁸[28] Evidence suggests that religion operates through two powerful channels: culturally it raises the desired number of children,²⁹[29] but it also creates supportive communities that encourage families and facilitate childrearing.³⁰[30] The cultural impact is so great that it appears to trump economic incentives. Not only do religious fertility differences cut across income lines, but a study of Eastern European fertility subsidies found that positive effects were limited to women from religious backgrounds.³¹[31] Again, secularism has the opposite effects, suppressing even religious minorities’ fertility within secular countries,³²[32] and overriding economic incentives to have children.³³[33]

Divorce

Religiosity, by culturally reinforcing marriage and family stability, may also reduce the negative effects of lower-cost, no-fault divorce.³⁴[34] A common view among sociologists is that rising female employment facilitated divorce through empowerment, but the evidence points to the reverse: the falling cost of divorce increases married women’s sense of economic insecurity and encourages them into the workforce.³⁵[35] Regardless, the associated increase in marital instability has had a clear negative effect on fertility rates.³⁶[36]

State pensions

The third unmentionable-but-well-documented driver of falling fertility is the rise of state pensions. Children were the original pensions, taking over the family farm or business and caring for parents in their old age. That relationship continued even in urbanized and industrialized societies. But state pensions “emancipated” parents from their own children. The elderly could now rely on the state for old-age support. Yet that support still comes from the next generation through taxation. This creates a serious free-rider problem. If you can rely on others’ children to fund your retirement, you don’t need your own. One study found that the introduction of state pensions in Brazil reduced female fertility rates by 1.3 children.³⁷[37] Studies of similar “natural experiments” across both industrialized³⁸[38] and developing countries³⁹[39] confirm large negative effects on fertility from the introduction of pensions.

Honest policy discussion

To be clear, I am not advocating unwinding of state pensions and modern divorce law, or encouraging everyone to get religion. But misdiagnosis of the causes of falling fertility is costly. We are richer, healthier, better housed, and better educated than any generation in human history, yet have fewer children. Demographic implosion has serious political and economic implications that deserve our attention. Hence, we need to be honest about why young people are choosing to have fewer children despite their measurably better living standards, and design solutions, however unpopular, that address the real problems.