The economics of Korea's disappearing small towns
More than half of Korea's municipalities score as extinction-risk on the index the country uses to track depopulation, and the policy response runs directly into the economics of agglomeration.
South Korea measures regional decline with an index borrowed from Japan: the number of women aged 20 to 39 in a district divided by the number of residents aged 65 and over. Below 0.5, a locality is classified as at risk of extinction, on the reasoning that a place without young women of childbearing age has no demographic mechanism for renewal regardless of what else it has. Korea Employment Information Service analyses using the measure have found well over half of the country’s 228 basic municipalities in the at-risk category by the early 2020s, up from around a third a decade earlier.
The demographic arithmetic behind that is national but its effects are local. Korea’s total population peaked at just under 51.9 million in 2020 and has declined since, with births falling below 250,000 a year. Yet the capital region — Seoul, Incheon and Gyeonggi province — passed 50 percent of the national population around 2019 and has kept growing, on a landmass of roughly a tenth of the country. Decline and concentration are happening simultaneously, which means provincial areas lose population faster than the national total falls.
The mechanism is a cascade rather than a drift. Young people leave for universities and jobs concentrated in the capital region and a few metropolitan cities; the remaining population ages; births fall further; schools consolidate and then close; the closure of a school removes the reason for families to stay, accelerating the next round. Health care follows the same logic — the health ministry designates medically underserved areas where residents cannot reach obstetric or emergency care within a reasonable time, and their number has grown. Bus routes, bank branches and grocery retail thin out along the same path, each departure raising the cost of remaining.
The fiscal system partly compensates. Local governments in Korea are heavily dependent on central transfers, and the shared local tax mechanism redistributes toward smaller, poorer municipalities, which keeps basic administration and services funded regardless of local revenue. Dedicated instruments have been added: 89 areas were designated as population-decline regions in 2021, a local extinction response fund of roughly one trillion won a year was established from 2022 for a ten-year period, and a hometown donation scheme launched in 2023 lets taxpayers direct contributions to regions other than where they live in exchange for tax credits and local goods.
Whether any of this can work is genuinely contested among economists. The agglomeration literature holds that productivity rises with density — through labour market matching, knowledge spillovers and supplier proximity — which implies that subsidising dispersal fights an economic force rather than a policy failure. On that reading, transfers can guarantee decent services in declining regions but cannot manufacture the economic base that would keep young workers there, and spreading small sums across many municipalities produces neither. The competing view emphasises that Korean concentration is itself policy-made — through the location of universities, public institutions and corporate headquarters — and is therefore reversible by the same means, as the relocation of public agencies to provincial innovation cities from the late 2000s attempted.
Recent policy has drifted toward a third framing that concedes the population will not return. The “living population” concept, given a legal basis in 2023, counts commuters, second-home owners and regular visitors alongside registered residents, on the argument that a town of 20,000 residents with 60,000 monthly visitors is a different economic proposition from one without them. Regional visa pilots aimed at attracting foreign workers and students to specific provinces work along similar lines, treating population as something to be circulated rather than only retained.
What is not in dispute is the timeline. The cohorts that would have to reverse provincial decline have already been born, in numbers that make reversal arithmetically impossible for at least a generation. The realistic policy question is which settlements are consolidated deliberately, and which are simply allowed to empty.