Geographic Concentration of Unsold and Completed-Unsold Housing in South Korea, 2025–2026

Overview This dataset provides a reproducible analysis of the geographic concentration of unsold housing and completed-but-unsold housing across South Korea's 17 provincial-level administrative regions from January 2025 through September 2026. The study uses 21 consecutive monthly statistical releases from the Ministry of Land, Infrastructure and Transport (MOLIT), Republic of Korea. It compares regional concentration patterns using the Herfindahl–Hirschman Index (HHI), the Gini coefficient, and regional shares of unsold housing inventory. Key Findings Between January 2025 and September 2026: Total unsold housing decreased from 72,624 to 65,941 units, a reduction of 6,683 units. Completed-but-unsold housing increased from 22,872 to 27,549 units, an increase of 4,677 units. The completed-but-unsold share of total inventory increased from 31.49% to 41.78%. The HHI for total unsold housing increased from approximately 957.07 to a bounded range of 1,041.91–1,055.63. The HHI for completed-but-unsold housing decreased from approximately 837.95 to 795.25. These results indicate that the geographic concentration of total unsold housing and completed-but-unsold housing changed in different directions between the first and final observation months. Dataset Coverage and Methods The analytical dataset covers 21 consecutive months and 17 provincial-level administrative regions, comprising 357 region-month observations. Regional concentration is measured using: HHI, calculated as 10,000 multiplied by the sum of squared regional inventory shares. The Gini coefficient, calculated across regional inventory counts. Regional distribution and completed-but-unsold inventory shares. The research package includes derived analytical tables, methodological documentation, data-quality checks, sensitivity analyses, and reproducibility code. Data Quality and Limitations The original statistical releases contain documented inconsistencies and aggregation limitations. In December 2025, two tables within the source workbook reported different completed-but-unsold inventory totals, differing by 414 units in Chungcheongnam-do. In July–September 2026, some tables aggregated Gwangju and Jeollanam-do. Where the separate regional counts could not be verified, concentration measures are reported as bounded estimates rather than exact values. The study does not establish causal relationships between housing policies and unsold inventory. Regional concentration measures do not directly represent housing market risk, vacancy rates, or unsold housing relative to the regional housing stock. The dataset extends preliminary regional unsold-housing analysis by providing a continuous 21-month series, reproducible concentration measures, and documented sensitivity tests. Original Data Source Ministry of Land, Infrastructure and Transport (MOLIT), Republic of Korea. Monthly Unsold Housing Statistics (미분양주택현황보고), January 2025–September 2026. Official statistics portal:https://stat.molit.go.kr/portal/cate/statMetaView.do?hRsId=32 Attribution and Reuse Original analytical materials were prepared by AptToSell. The Creative Commons Attribution 4.0 International (CC BY 4.0) license applies to the author's original analysis, derived tables, documentation, and code. Original government source materials retain their applicable source-specific usage conditions. The dataset is intended to support reproducible research, housing market analysis, journalism, and public understanding of regional unsold housing patterns in South Korea.