The vertical density of urban apartments is catastrophic for fertility
The gentle density of suburban subdivisions is much better.
Written by Daniel Hess.
I have often argued that urban density is a problem for fertility. In a thoughtful piece, Noah Carl and Bo Winegard challenged me by arguing that population density is only weakly associated with fertility rates across US states and EU member states. They maintain that the scope for boosting fertility by reducing density is limited, and claim that the low fertility rates in dense urban centres are largely explained by selection. Carl and Winegard suggest that I haven’t done enough to show causality, though they do say, “it is conceivable that the very high population densities of the kind seen in Chinese city states are detrimental to fertility.”
Related to their argument is Bryan Caplan’s claim – from his graphic novel Build Baby Build – that there are big economic gains to be had by throwing out zoning rules and allowing unlimited vertical building. I have responded to Caplan with numerous examples of places that built like this and saw fertility plunge to catastrophically low levels. Caplan then retorted that the very low fertility of urban cores is really due to high housing prices and if you could just build even more high-rise apartments, they would become cheap, and fertility would recover.
My responses to these arguments can be summarise as follows:
Suburban subdivisions with single-family homes are quite dense compared to rural areas, but they don’t seem to harm fertility.
High rise apartment towers (as they are usually built these days) are catastrophic for birth rates, and the causes are clear: there is no yard for kids to play in; kids bother the neighbours; and most apartments are too small for families.
Caplan’s vision of unlimited vertical building has already been tried in dozens of Asian cities, and the results have been catastrophic. Fertility rates are 0.7 or lower in places like Seoul, Shanghai, Beijing and Bangkok – even though Asian apartments are incredibly cheap by western standards, thanks to their abundance.
Carl and Winegard estimate that limiting Britain’s density to that of Suburban England would net 0.18 births per woman in fertility and say, “this is hardly going to solve the fertility crisis”. Yet combined with other measures, a difference of 0.18 is a lot. Most pro-natal policies in Europe yield much weaker effects.
As to selection, there is no such effect in China or Korea because almost all homes are apartments in high-rise buildings. Yet these places still have very low fertility.
How dense is too dense?
I agree with Carl and Winegard that national-level measures of density are not great. The density where I live, in Montgomery County Maryland, is fairly high at 808 people per square kilometer. The density of neighboring West Virginia is low – just 28 people per square kilometer. Yet the fertility of both is around 1.6, comparable to the national average.
What’s going on? The majority of Montgomery County, Maryland is suburban with spacious single-family homes, guided over the years by zoning rules that reinforced family-friendly houses.
Meanwhile, the area adjacent to DC, where apartments in high rises are much more common, has a TFR of around 1.1 (0.9 for non-Hispanic whites). Both areas are politically liberal, highly educated and have incomes well above the national average. Fertility-wise, DC is similar to East Asian countries, while Montgomery County is comparable to France. A vast difference.
This illustrates well the threshold where density and crowding become a problem for fertility. Apartments in towers just aren't made for families.
According to modern norms, a two-child family with one boy and one girl requires three bedrooms because the parents need their own bedroom and opposite-sex kids above a certain age also need their own rooms. (Child Protective Services does not look favourably on families that don’t follow this.) But just seven percent of new apartments in the US have three bedrooms.
So, rising density isn’t much of a problem going from rural West Virginia to suburban Maryland. But it becomes a big problem when people are squeezed into small apartments that are ill-suited for families. Carl and Winegard’s density analysis (at the level of states and countries) can't see this because density and crowding are local.
However, their critique forces me to do better than handwave. What we see in detailed geographic maps is that fertility falls off a cliff in dense urban centres. Consider Sydney and Perth in Australia, where the TFR is a healthy 2.0 in the expensive but family-friendly suburbs but is an ultra-low 1.0 in the apartment-centric core.
Suburbs are a wonderful answer to the terrible fertility problem of cities. Workers are networked into the productive urban economy but also have housing that is suitable for families. Asia embraced the first part of economic development (urbanization) but missed the second (suburbs), and now they are in big trouble.
We should not forget the lessons of the Baby Boom. The economy grew rapidly at the same time as it was de-densifying! There was an explosion of suburban building that got people out of the low-fertility cities! From Green and Wachter (2005):
With the strong expansion of the U.S. economy in the post–World War II period driving up incomes, together with the new institution of the long-term (and therefore affordable), fixed-rate, self-amortizing mortgage, homeownership expanded rapidly. America was transformed from a nation of urban renters to suburban homeowners.
A high rate of marriage was central to the Baby Boom but so was suburban living. Those two pillars together supported the rise in fertility in the US, as well as many other countries.
Egypt and other drivers of fertility
Carl and Winegard offer the example of Egypt, which is incredibly dense along the river Nile, and yet has a TFR of 2.8, well above replacement.
But Egypt is a poor country with a high level of religiosity. Its GDP per capita is just $3,500, and Pew Research reports that more Egyptians believe Sharia law is the “revealed word of God” than do Afghanis. This isn’t the prescription we want, is it? And unlike in the United States, there is no norm in Egypt that says you can't put a large family in a one or two-bedroom apartment.
Monaco, the exception that proves the rule
Carl and Winegard point to Monaco as an example of a place in Europe that is both dense and has high fertility. Europe's densest country, it lists a fertility rate above 2.0! Alas, Monaco reports its fertility incorrectly. They count every child born in the country but often not the parents. This is a recipe for wildly overstating fertility, especially since many people intentionally seek to get their kids highly sought-after Monegasque citizenship.
Is the effect of density causal?
Demographer Lyman Stone, who has pored over fertility studies more than anyone else I know, offered these thoughts on the relationship between density and fertility:
The best evidence we have suggests that living in dense environments causes lower fertility. We know that selective migration does not account for most of the suburban-urban fertility difference. We also know that price/fertility gradients are steeper in denser environments.
We also know that in non-human species reproductive success declines with environmental density even when food stress does not occur. We also know that exogenous shocks to house size or house quality have expected effects on fertility.
In general, housing/marriage/location/baby are problematically confounded decisions; they all go together. So, studying this is all a bit tricky. But there are basically zero studies that have been able to argue that density was pro-natal, on any measure of density.
Now, there remains debate about why density is bad for babies. Some people argue density qua density is bad: basically, they think humans face an evolutionarily crowded environment in cities which triggers serious status anxiety and thus suppressed reproduction.
Personally, I think there's decent evidence for both of these lines of thinking. Density probably does provoke certain intensified anxieties and also it really is hard to deliver family-optimal housing in dense environments.
Why housing matters so much
Carl and Winegard show that, at the national level or the level of US states, population is barely related to fertility. I have hopefully shown that this is the wrong measure to look at.
We need to zoom in to see what kind of housing people actually live in. And when we do that, we find that the greater the share of a city’s or country’s housing is apartments, the lower its fertility tends to be.
The abundance of affordable but family-unfriendly apartments across Asia has done nothing to help fertility, and seems to have made things a lot worse.
For one thing, cheap apartments make it much easier to stay single, thereby reducing people’s motivation to couple-up. For another, apartments lead to very small family sizes. Only around 5% of births in Seoul are a third birth and only 0.6% are a fourth birth or more. The percentages in the US are 30% and 13%, respectively!
When you combine these two factors – more singles and tiny families – you end up with the impossibly low fertility levels we see in Asian cities of 0.5 to 0.7 births per woman. Civilization cannot survive long with fertility rates that low.
But we have a ready solution! Suburbs offer the economic benefits of urbanization without the fertility costs. In a world that is demographically collapsing, it’s imperative that countries build the housing typical of America and Australia, and avoid the catastrophic path of East Asia.
There are two important reasons to focus on housing:
Housing is forever. Most houses and apartments will far outlast us. Housing built today will affect fertility in a positive or negative way for generations to come. What else can we do that will affect fertility on such a long time horizon? Tragically, many countries have mainly built an anti-natal type of housing, and they will be stuck with that housing for many decades. Avoid that if you want to have a future!
Housing is something policymakers can actually affect. You can’t easily make people religious or legislate that they fall in love. But you can have rules requiring housing to be family friendly. (In fact, we already have rules favoring single family homes – just don't throw them out the window.)
What can be done with apartments that have already been built?
Cities like Seoul, Tokyo and Shanghai are filled with high-rise apartment towers, which contributes to their incredibly low fertility. Are they stuck forever with housing stock that isn’t family friendly?
One solution is to combine small apartments to form larger ones. This is much easier than converting, say, an office tower into residential housing. Apartments with one or two bedrooms can be combined to create units with four or five bedrooms, which would be much more suitable for families. In the future, there will be a surplus of apartments in many cities, as countries depopulate. Combining these into larger units will help deal with this surplus while also helping to solve the fertility crisis. Although large apartments are not ideal for families, they would be much better than the small apartments that currently dominate.
In conclusion: Build, baby, build! But outward, not upward!
A slightly different version of this article was originally published on X.
Daniel Hess is a researcher, writer and pro-natal activist. You can find him on X at More Births.
Our Comments
Written by Noah Carl and Bo Winegard.
We are grateful to Daniel Hess for taking the time to respond to our article criticising his arguments about density and fertility. While his points are well-taken, we remain unconvinced that population density (or housing type) is a major cause of low birth rates.
In making his case, Hess puts considerable emphasis on the fact that fertility rates tend to be low in dense urban centres where many people live in apartment towers. He interprets this as evidence that building such towers causes people to have fewer children. Our interpretation is different: we suspect that such towers are a consequence of people’s decisions to have fewer children.
There is no inherent reason why apartments can’t have three bedrooms or thick walls that muffle the sound of babies crying. Did developers wake up one day and just decide to build apartments that aren’t family friendly? No, they are responding to demand. People who aren’t interested in having large families do not demand family-friendly apartments. (Note that in terms of square feet, US homes are actually larger now than they were in the early 20th century, when fertility was higher.)
Why aren’t people interested in having large families? For all the usual reasons: rising affluence, declining religiosity, the higher status of women, the collapse of marriage. In other words: these factors cause people to want fewer children, and people’s desire for fewer children lowers demand for family-friendly apartments.
Of course, not everyone is affected to the same degree by fertility-suppressing trends, which is why not everyone ends up with the same number of children. And it so happens that the sorts of people who are more affected (liberals, atheists, career women) prefer living in dense urban centres. This leads to an association between urban density and low fertility that we believe is much stronger than the causal effect.
Regarding Britain, we consider 0.18 additional births per woman to be unimpressive given the near-impossible reduction in density it would require (demolishing most of the buildings in central London and every other major city). While the impact might be large relative to other interventions that have been tried, the costs would be gigantic.
Regarding Egypt, we were not putting it forward as a model for Western society. We only mentioned it to illustrate the point that unweighted density is an unsatisfactory measure in this context.
Regarding Monaco, Hess points out that its TFR is overstated, linking to a tweet by Twitter-user Birth Gauge. Yet Birth Gauge suggests that its true TFR is “likely similar to the level of France (around 1.8)”. This would mean that Monaco still has a higher TFR than Italy despite being six times denser.
Hess is right that housing matters insofar as whatever gets built now will last for decades. But it’s no good building single-family homes or four-bedroom apartments if no one wants to live in them. You first have to change the culture.
Noah Carl and Bo Winegard are the Editors of Aporia.
Support Aporia with a paid subscription:
You can also follow us on Twitter.
"I have often argued that urban density is a problem for fertility."
Not to mention sanity. I fail to see an advantage to living in a mega-city. They no longer produce anything of use for the country as a whole. The economics consists of financial products and the service industry. If mega-cities disappeared overnight, the rest of the country would have little notice.
"Caplan then retorted that the very low fertility of urban cores is really due to high housing prices and if you could just build even more high-rise apartments, they would become cheap, and fertility would recover."
I reply that only an idiot would want to live in a place like that.
"Both areas are politically liberal, highly educated and have incomes well above the national average."
The term 'highly educated' is so nebulous. What is left unanswered is educated in what?
Furthermore, they have incomes well above the national average only because the cost of living is well above the national average, which means no advantage.
"A high rate of marriage was central to the Baby Boom but so was suburban living. Those two pillars together supported the rise in fertility in the US, as well as many other countries."
Excellent point.
Now, my reply to the rebuttal:
"Did developers wake up one day and just decide to build apartments that aren’t family friendly? No, they are responding to demand."
No, they are responding to more profit.
"Note that in terms of square feet, US homes are actually larger now than they were in the early 20th century, when fertility was higher."
Homes, yes; apartments, not at all. Many newer NYC apartments have been declining in size for many years.
"And it so happens that the sorts of people who are more affected (liberals, atheists, career women) prefer living in dense urban centres."
There you go again with the degradation of atheists. Why the hell would atheists prefer dense urban centers? And why would they shun children?
And, as always, quality over quantity.
People in cities also often only live there temporarily, moving to the suburbs to have kids. So the low fertility of urban areas is not just inflated by the type of people who choose to live there, but also by “fertility drain” of urban residents