KIRD collects and reorganizes the official statistics that count immigrants living in Korea. It is not a government publication, and every source table is published by Ministry of Justice and Ministry of the Interior and Safety and Korean Educational Development Institute and Ministry of Unification.
Statistical Yearbook on Immigration and Foreign Policy and Status of Foreign Residents in Local Governments are public, but each ministry counts a different population, so comparing them directly gives mismatched figures. Registered foreigners, staying foreigners, and foreign residents have similar names and very different counts. Deciding which figure to cite takes time, and at the district level even finding the data is difficult.
Linking one year to the next is not simple either. Each yearbook is issued separately, and the table layout, the nationality names, and the status-of-stay codes differ from edition to edition, so a district in 2009 does not line up with the same district in 2019. Administrative boundaries also change over that period. Incheon Nam-gu became Michuhol-gu in 2018, Cheongwon-gun was absorbed into Cheongju-si Cheongwon-gu in 2014, Yeongi-gun became Sejong in 2012, and Gunwi-gun moved from Gyeongsangbuk-do to Daegu in 2023. Left as they are, the names break the time series for one area partway through.
Comparison across districts also requires the denominator and the index formula to be the same in every year. From 2014 the yearbook lists every nationality at the district level, so a diversity index computed on the data as published jumps in that year even when the population does not change. The dataset was built to remove that step.
The work of linking yearbooks whose structure changes every year is done by the public pipeline. The same code produced the tables, which can be downloaded from openICPSR and recalculated yourself.
| Section | Contents | Selectable units |
|---|---|---|
| Registered foreigners | By district, by nationality, by language, by status of stay | 250 districts, 192 countries, 1,361 languages, 35 statuses |
| Population groups | International students, multicultural families, nationality acquisition, refugees, North Korean defectors | Five populations |
| National map | Registered foreigner share by district | 250 districts, by year |
| Service map | Organization list, service gaps | 903 organizations, 244 districts |
| Guides | Health coverage, grant programs | Treatment cost paths and 41 grant programs |
| Data and about | Open dataset, indicators and scope, about | 17 files, 1,166,660 rows |
Source: The numbers in this table were counted from the data files the pages actually read.
Every number in the tables and figures comes from the open dataset. It holds seventeen tables that merge the MOJ and MOIS yearbooks onto one standard, with year, region, nationality, and status of stay linked on the same basis.
Each yearbook is issued separately, and the table layout, the nationality names, the status-of-stay codes, and the district names differ from edition to edition. Left as they are, a district in 2009 does not link to the same district in 2019. The pipeline standardizes the notation across editions and applies the administrative reorganizations so that the years connect.
Shares, diversity, and segregation indices are calculated the same way in every year. Shannon diversity uses the top 19 nationalities of each year plus a remainder. The yearbook has listed every nationality at the district level since 2014, and using all of them would move the index sharply in 2014 even when the population did not change.
MOJ registered foreigners are foreign nationals staying longer than 90 days, and MOIS foreign residents add people who have acquired Korean nationality and their children. Adding the two counts the same person twice. Nationally, foreign residents in 2008 were 1.05 times registered foreigners and in 2024 1.74 times. Both the level and the trend change with the definition.
Source: openICPSR open dataset, Pipeline repository README.
Note: Row counts and year ranges by table are in Open dataset page.
Every page opens with a paragraph stating the size and composition of the selected unit in sentences. Values fill a fixed sentence template, so when the unit or the year changes, only the values change in the same positions.
Every item mentioned in the summary paragraph has a numbered table or figure. View as table under each figure holds the exact values that figure used.
Every plate carries its source in the source line and its method and caveats in the note line. Which table from which ministry a value came from is visible inside the plate.
When you cite a figure, state which basis it uses. The differences between the bases are set out in Indicators and scope.
Note: Tell us about errors in the data or missing organizations and we will fix them. This site reorganizes published statistics and is not a government publication.
Yoo, N. (2026). Spatiotemporal administrative dataset of nationality, residential diversity, and language demand among immigrants in South Korea, 2006–2024 [Data set]. Ann Arbor, MI: openICPSR. https://doi.org/10.3886/E249944V1
The open dataset is CC BY 4.0. Copyright in the source data belongs to each ministry.