42 economies shrank in 2024 — and 17 of them recorded more births than deaths. Fertility does not cleanly sort the declines; the pattern is consistent with net outward movement

The world added people in 2024 — its population grew 0.96%. Not everywhere. We pulled the World Bank's population-growth series (SP.POP.GROW, 2024 revision, updated 13 July 2026), which carries the United Nations' latest estimates, and kept every non-aggregate economy with a 2024 value: 217 of them. In 42, the population fell. Those 42 hold about 2.0 billion people. The decline in most of them is told as a birth story — a fertility rate below replacement, a maternity ward gone quiet. Read all 42 against their own vital statistics and the birth rate does not sort them. In 17 of the 42, more people were born than died in 2023; their numbers fell anyway. Seven carry a fertility rate at or above the 2.1-children-per-woman replacement line and are shrinking regardless. The clean cases are large enough to make the pattern visible, though these data do not quantify the uncertainty around the residual. Nepal recorded 12.4 more births than deaths per 1,000 people in 2023 — a firmly growing natural base — and still lost population. Sri Lanka (+3.0 per 1,000 of natural increase), Albania (+1.9), Jamaica (+3.5) and Mauritius (+0.8) did the same: births beat deaths, the population shrank. The gap is a residual — 2024 growth minus 2023 natural increase, both as percentages (the crude birth-minus-death rate divided by ten) — that is consistent with net outward movement, though it also absorbs census rebenchmarking, the one-year offset between those two vintages, and is not a joined migration count; read the magnitude as modeled and the direction as plain. Where births beat deaths yet the population fell, that residual is consistent with a net outflow. ^mmodeled Now invert the list. The economies with the world's lowest fertility sit at the shallow end of the decline, not the steep one. Hong Kong, at 0.75 children per woman the lowest here, shrank 0.16%. China, at 1.00, shrank 0.12% — still the largest absolute fall, roughly 1.7 million people. Italy (1.21) and Thailand (1.21) each shrank 0.05%. The four steepest declines — Kosovo (−5.39%), St. Martin (−5.17%), the Marshall Islands (−3.35%) and Moldova (−2.28%) — carried more births than deaths in three of the four.

The following is analysis, not fact. The depopulation story most desks tell is a maternity-ward story: a birth strike, baby bonuses, a fertility rate under replacement. It names a real long-run force. But it does not cleanly sort which of these 42 economies shrank in 2024, or how fast. Among the 27 shrinkers with more than a million people, our calculation puts the correlation between 2023 fertility and 2024 population growth at −0.20 — a weak association, and tilted the wrong way for the birth-rate story: higher fertility is not associated with a slower decline in this sample, if anything the reverse. The ranking is not cleanly sorted by fertility; the remaining change is consistent with migration contributing to some of it, though these data do not establish its size. Kosovo's maternity wards are busy; its population still falls. Nepal births far more than it buries and its numbers still drop — consistent with a sustained net outflow, though the 2024 residual shows only that net loss, not where it went. The pattern is consistent with migration contributing to these 2024 declines, its size not established here, with fertility the longer-run force; migration barely enters the natalist policy debate that population decline usually sets off. Why do the famous low-fertility giants — China, Italy, Japan, Thailand — sit so shallow? Momentum. Decades of higher birth rates left them with large adult cohorts still alive and not yet at the death-heavy ages, so a 1.0 or 1.2 fertility rate has not yet arrived in the death column. Their decline is loaded, not fired. Read the two halves of the list together and the honest reading is a split screen: the migration-consistent residual tracks the 2024 ranking, while low fertility is the more plausible force behind the longer-run trajectory that ranking will eventually bend toward. That momentum reading is consistent with these countries' age structures, but we did not decompose it here — no cohort model was run — so treat it as a hypothesis the pattern supports, not a mechanism these data isolate. ^sspeculative

The strongest case against this piece is that one year's ranking is not destiny. China shrinking 0.12% and Italy 0.05% looks mild only because their past-born cohorts are still alive; a fertility rate near 1.0–1.2 makes a steeper natural decrease increasingly plausible as those cohorts age into the death-heavy years. On that reading fertility is the master variable everywhere — merely invisible in the giants for now — and migration the cyclical layer on top. It is a fair account of the same numbers, and the long-run demographic literature backs it. Two caveats also bound the finding: the extreme small-state and Kosovo magnitudes can reflect census re-benchmarking rather than one year's real flow (the count-based facts — births versus deaths, above-replacement fertility — do not depend on those magnitudes), and the growth figures are 2024 while the vital rates are 2023, a one-year offset. ^sspeculative

**View from a shrinking capital in the Balkans or the South Pacific.** To a statistics office in Pristina, Chisinau or Majuro, the Western debate over baby bonuses and fertility clinics can look like the wrong diagnosis. Several of these places have fertility at or above replacement and still shrink, so a natalist subsidy is not obviously the lever. What the residual is consistent with is people leaving on net — which birth-rate policy is not designed to reach. Scored on fertility, they look unexceptional; scored on who is still there next year, they are among the places losing population fastest.