Does an Expensive Childhood Mean Fewer Children? Weaker Than the Story Suggests
Across 61 countries, childhood cost and fertility correlate at −0.56 — real, but the weakest relationship on this site, and mostly income in disguise.
Across the 61 countries with a published cost figure, the cost of raising a child and the fertility rate correlate at −0.56 — a real relationship but the weakest between any two indicators on this site, and one largely explained by national income.
The intuition is strong: children cost more, so people have fewer of them. The data supports it and supports it less firmly than almost anything else on this site.
What the numbers are
| Most expensive | Cost to 18 | Fertility |
|---|---|---|
| United States | $310,000 | 1.62 |
| Australia | $300,000 | 1.60 |
| Norway | $290,000 | 1.40 |
| Canada | $280,000 | 1.30 |
| Least expensive | Cost to 18 | Fertility |
|---|---|---|
| DR Congo | $7,000 | 6.0 |
| Mozambique | $8,000 | 4.6 |
| Ethiopia | $9,000 | 3.9 |
| Uganda | $9,000 | 4.4 |
A forty-four fold difference in cost, and roughly four times the fertility at the cheap end. Put like that the case looks closed.
The correlation, measured
Across all 61 countries with a published cost figure: r = −0.56.
That is a real negative relationship. It is also the weakest correlation between any two indicators on this site. For comparison:
| Pair | r |
|---|---|
| Under-15 share × fertility | 0.95 |
| Life expectancy × fertility | −0.83 |
| Secondary completion × fertility | −0.75 |
| Cost of childhood × fertility | −0.56 |
If cost were the mechanism people assume, it should sit near the top of that table rather than the bottom. Schooling predicts fertility better than price does.
Why it is weaker than it looks
Both are tracking income. A country where a childhood costs $310,000 is a rich country, and rich countries have lower fertility for a long list of reasons that arrive together — education, urbanisation, healthcare, women's employment, later marriage. Nearly every indicator in this dataset correlates because they are all measuring development stage, and cost is one more of them.
The cost figures measure different things. Some national estimates include housing and childcare, some exclude both, most stop at 18. A forty-four fold range across methodologies that do not agree is a noisy variable by construction.
And the direction is not established. A correlation across country averages cannot tell you whether expense suppresses births, whether smaller families spend more per child, or whether a third thing produces both. All three are plausible and this data separates none of them.
What this does support
The second child is not priced like the first. Housing, equipment and most fixed costs are already paid, so per-child averages mislead in both directions — about 32% of the US total is spent before school age, and that share falls sharply for a second.
Cost is a real constraint on timing. Even where it does not change how many children people have, it changes when — and later first births compress the window in which more are possible. That is a mechanism the correlation above cannot see, because it compares totals rather than sequences.
Nothing here is advice about your household. A national average is not a budget, and the range within any country dwarfs the range between the ones in the table.
See the full cost table at cost of raising a child by country, what it does and does not include at how much does it cost to raise a child, or the wider question at why is the birth rate falling.