Korea Ran the World's Most Expensive Test of Whether a State Can Buy Births
Two decades of subsidies bought a 0.72 fertility rate and worsening maternal mortality — the 0.99 rebound is being claimed as vindication, but the timing indicts the money, not acquits it.
South Korea just produced the number everyone watching population decline has been waiting for: a fertility rebound, 0.99 in 2025, up from the 0.72 that marked the deepest low any country has recorded. Cue the vindication tour — the subsidies worked, the spending finally bit, copy the model before your own pyramid inverts. Before anyone copies it, look at what preceded the rebound: roughly two decades of pronatalist spending that, summed, runs to about the GDP of a mid-sized country, deployed against a birth rate that fell anyway (The Spring Group). The strongest version of the vindication claim is simple and worth granting: maybe the cash was never wrong, just slow, and cumulative — a price that finally crossed some household's threshold. It's a real possibility. It's also, on the timing, almost certainly not what happened.
Start with what the spending was actually *for*, because the record answers that question more honestly than any policy paper. The alarming part isn't the wasted money. It's that a state can spend the GDP of a mid-sized country trying to acquire births and still let maternal mortality worsen while it does. That tells you the object of policy was births — the count, the output, the statistic — not the people having them. And a policy built to acquire an output rather than protect the people producing it will do exactly what this one did: price the transaction and ignore the constraint.
A state that spends a country's GDP acquiring births while the people having them die more often wasn't failing at natalism. It was succeeding at something else.
The constraint is marriage, and Korea is an unusually clean place to see it. Births outside marriage run around 2% there, so essentially all fertility passes through the marriage gate — which makes the last few years the closest thing to a natural experiment anyone has on "norms, not cash." Marriages and births began moving together from mid-2022, before any new subsidy could plausibly have bitten. The curve bent precisely where the binding constraint sat, not where the money sat. If the cash were the mechanism, you'd expect the bend where the cash landed. It bent at the gate instead (Georgetown Journal of International Affairs).
So here is my bet, stated so it can lose: the 0.99 is pandemic postponement landing — weddings deferred across 2020–22 converting into births on a lag — a one-time catch-up, not a level shift. Quarterly numbers can't settle that; the annual series can. What would change my mind is simple: annual figures that hold above the old trend for years, with no marriage-spike residue left to explain them. If that shows up, the subsidies — or something durable that arrived with them — earned the credit and I'll say so. Until then, a rebound that begins before your intervention could plausibly act is not your intervention working. It's your intervention getting lucky with its press release (Think Global Health).
The general lesson is the uncomfortable one, and Korea hasn't answered it — it has only just started testing it. You cannot buy your way out of a norms problem. You have to change the norms, and nobody yet knows whether a state can do that on purpose, or how long it takes when it can. Every aging economy on the planet is watching this experiment, and most of them are hoping the answer is cash, because cash is the instrument they already have. The Korean record so far suggests they're hoping for the wrong thing (NCFR, Bogenschneider & Lee).