Urban fertility is markedly lower than rural fertility. In the U.S., for example, rural women ultimately have about .3 more children than urban women. The obvious story is that (a) urban housing is more expensive, so (b) urban women live in smaller homes, which (c) makes having kids less pleasant. Raising my four children in a two-bedroom DC apartment would have meant much more stress and much less sleep than raising them in my suburban McMansion.
On second thought, though, perhaps this obvious story is wrong. Maybe the real story is selection. People who don’t want many kids choose to live in cities and enjoy their many child-unfriendly amenities. People who want many kids, in contrast, make the opposite choice. Popular stereotypes tell us that couples routinely “move to the suburbs to raise a family.” Could this explain the whole fertility pattern?
There are many empirical strategies you could use to approach this issue, but here’s one that’s clean and simple: Look at fertility in city-states. Unlike people in virtually every other country on Earth, the citizens of city-states have greatly restricted opportunities to “move to the suburbs to raise a family.” So how does their child-bearing respond?
The sample is admittedly small: Officially, there are only Singapore, Monaco, and Vatican City. We should throw out the Vatican for obvious reasons. But given their history and “One Country, Two Systems”, Hong Kong and Macau should count, too. Results:
City-State Total Fertility Rate
Hong Kong 1.19
There are very low rates, but how do they compare to those in regular countries with similar cultures? The TFR in mainland China is 1.60, far above the rates for Hong Kong, Macau, or Singapore. You could argue, though, that highly-developed Taiwan is the better comparison country. Its TFR is now 1.13 – comparable to Hong Kong, but still well above Macau or Singapore. The obvious comparison country for Monaco is France, with a TFR of 2.07.
Overall, then, I say city-states credibly support the (high housing costs –> small homes –> few children) causal chain. If you know of better evidence for or against, please share in the comments.