Marriage Migrant Status Brief

Marriage Migrants in Korea: Naturalization, Permanent Residence, and Visa Status

Brief · November 9, 2026 · Sources: statutory texts (law.go.kr), KOSIS and Immigration Statistics Yearbook, National Survey of Multicultural Families 2015, 2018, 2021, 2024

Summary

Using the data and analysis in this report, we can describe, at the macro level and over time, which residence and status options the 1998–2011 reforms opened to people without Korean nationality, and where each group ended up. We cannot estimate reform-by-reform causal effects or analyze at the individual level who made which decision. One such claim we got close to was that allowing multiple citizenship increased the likelihood of nationality reinstatement among overseas Koreans (mostly Korean Americans), but the effect is significant only at the 10% level (randomization inference p = .07 one-sided, .10 two-sided).

A microanalysis we can do is an estimation of which factors are correlated with an individual's preference to naturalize. Origin matters far more than any other factor. Among individual characteristics, college education (−5 percentage points in wanting to naturalize) and having met one's spouse without an intermediary (−6 points) are associated with a lower preference, while fluent Korean (+3 points) is associated with a slightly higher one. Household income, children and urban or rural residence matter little. The origin differences remain after accounting for years in Korea, sex, Korean proficiency, own education, household income, marital status, children, how the couple met, and urban or rural residence. Korean proficiency and children are strongly associated with whether people actually naturalize (about +9 and +8 points), but only weakly with whether they want to.

Vietnamese women mostly naturalized. Mainland Chinese women took permanent residence more often than the other three main groups, and Korean-Chinese marriage migrants mostly naturalized, with about a third of the non-naturalized holding permanent residence. Permanent residence therefore partly explains why Chinese naturalization fell. It accounts for about a third of the drop between the 2000–04 and 2009–10 entry cohorts of mainland Chinese women, and in 2015 non-naturalized Chinese and Korean-Chinese women were 29 and 20 points more likely than Vietnamese women to want permanent residence only.

Filipinas neither naturalized nor took permanent residence at rates similar to Vietnamese or Chinese women; most remain on the marriage visa. Our hypothesis was that the way Filipinas married or their education made naturalization less attractive and so lowered their naturalization rate. The data do not support it. Filipinas are more often college-educated than Vietnamese women (26% vs 6%), and college education is associated with a slightly higher share not planning to naturalize. Yet after controlling for education, meeting route and the other factors above, the share of Filipinas with no plan to naturalize is only 2–3 percentage points higher than among Vietnamese women (about 6 points if "don't know" is counted as no plan), while their naturalization is 13 points lower. Most non-naturalized Filipinas plan to naturalize (58–82% across the 2005–14 entry cohorts and surveys; see "Status and plans of the same cohorts at each survey"). The Philippine gap lies in realizing naturalization, not in wanting it.

Hwang (2015) used the 2012 National Survey of Multicultural Families to relate naturalization to six origin-level factors: dual citizenship, size of the ethnic group, discrimination, geographic distance, and economic and political gaps with Korea. Each factor, however, is a dummy for a set of origin countries. "Dual citizenship" is coded 1 for Vietnam and the Philippines, "size of the ethnic group" for Chinese and Korean Chinese, and "economic gap" for Japan alone. With most marriage migrants coming from just four groups (mainland China, Korean China, Vietnam and the Philippines), these dummies cannot separate one origin-level factor from another; each coefficient is a contrast between particular nationalities. Dual citizenship is also a poor fit for these data. Korea allowed naturalizing marriage migrants to keep their original nationality only from January 2011, and by Hwang's own figures most marriage migrants who had naturalized by 2012 did so before 2011, when they had to renounce their original nationality. Our results point the same way: naturalization among Filipinas, whose origin country does not strip citizenship, fell after 2011, and we find no effect of the 2011 reform on marriage migrants' naturalization.

Estimates in this summary come from models of naturalization and naturalization plans in the 2015–2024 surveys (average marginal effects relative to Vietnam; full report Tables 9-4 and 9-5; scripts 79 and 80). The reinstatement estimate comes from a Poisson difference-in-differences across 15 origins (full report Table 10; script 69); Hwang (2015), Asian and Pacific Migration Journal 24(3): 376–402. Survey-time characteristics such as Korean proficiency, education and income may themselves follow naturalization, so these are associations, not causal effects.

Naturalization intent: model results

Average marginal effects in percentage points, each relative to the reference category in parentheses, holding all other variables in the table constant. Larger negative values in the first column mean less wanting to naturalize.

E1D1 (three outcomes)E3
Wants to naturalizeNaturalizedIntends, not yetNo planWants PR only (2015)
Origin (vs Vietnam)
Philippines−2.4***−12.9***+10.2***+2.7***+3.8*
Mainland China−14.3***−11.9***−2.5**+14.5***+28.9***
Korean-Chinese−4.4***−3.0**−2.3*+5.2***+20.0***
Taiwan, Hong Kong−9.4***−21.2***+11.0***+10.3***+16.7***
Thailand−4.6***−29.8***+23.1***+6.7***+14.8***
Mongolia−7.8***−22.4***+14.7***+7.7***+12.3***
Cambodia+0.6−2.0*+2.6**−0.6−1.5
Japan−55.6***−60.7***+0.6+60.1***+46.0***
Years in Korea (vs 6–9)
0–2 years−0.2−45.6***+41.4***+4.3**−1.4
3–5 years−0.9−27.1***+24.6***+2.5**−1.3
10–14 years+3.9***+17.1***−13.0***−4.1***+0.9
15–19 years+6.9***+28.3***−21.2***−7.0***−0.8
20+ years+8.2***+32.4***−23.7***−8.7***−4.5
Sex (vs male)
Female+1.9**+3.1**−1.4−1.7**−7.5**
Korean proficiency (vs average, 2.5–3.5)
Fluent (4.5+)+2.7***+8.9***−6.5***−2.4***−3.8*
Good (3.5–4.5)+1.1*+5.2***−4.1***−1.1*−0.3
Low (<2.5)−2.3***−13.1***+9.4***+3.6***−0.8
Own education (vs high school)
College or more−4.9***−4.1***−0.6+4.7***+6.4***
Middle school or less−0.7−3.9***+3.2***+0.7−0.6
Missing−10.2−14.9*+5.4+9.5–
Household income, KRW/month (vs 2–3 million)
Under 2 million−0.5+1.0−1.5*+0.5−1.1
3–4 million−0.7+1.5*−2.2**+0.7+0.9
4 million or more−1.8***+0.9−2.8***+1.8***+1.0
Marital status (vs married)
Divorced or separated+2.8+12.2*−11.1**−1.1+28.1
Widowed+3.5+8.0−5.0−2.9+6.2
Children (vs none)
Has children+0.9*+7.7***−6.8***−1.0**−2.8*
How the couple met (vs marriage agency)
Family or relatives−0.1+0.1−0.4+0.3+1.0
Friends or coworkers−2.0***−3.9***+1.7*+2.2***+1.3
Religious organization−2.0**−1.2+0.9+0.4−2.8
On their own−5.9***−9.1***+2.8***+6.3***+5.7*
Missing+1.3+4.7−3.4−1.3−10.7
Residence (vs urban dong)
Rural (eup/myeon)−0.1−1.4**+1.4**0.0−1.3
Survey year (vs 2018)
2015+1.9***+9.7***−7.4***−2.2***–
2021+1.6***+3.2***−1.7*−1.5**–
2024+2.9***+3.1**−0.4−2.7***–

E1: logit of wanting to naturalize (naturalized or plans to) vs no plan, 47,245 respondents in the 2015–2024 surveys, "don't know" excluded (90% want). D1: multinomial logit of naturalized / intends / no plan on the same sample (63% / 27% / 10%). E3: multinomial logit among 5,808 non-naturalized respondents in 2015, the only survey offering "PR only" (Korean nationality 71%, PR only 24%, neither 6%). All models control for years in Korea, sex, Korean proficiency, own education, household income, marital status, children, how the couple met, urban or rural residence and survey year; average marginal effects with robust standard errors. * p < .05, ** p < .01, *** p < .001 (two-sided, robust standard errors). E3 uses the 2015 survey only, so it has no survey-year terms. Counting "don't know" as no plan puts the Philippines at −5.6 in E1. Source: full report Tables 9-4 and 9-5; scripts 79 and 80.

The macro story

We can describe, at the macro level and over time, which residence and status options the 1998–2011 reforms opened to people without Korean nationality, and where each group ended up. We do not estimate reform-by-reform causal effects or who made which decision.

Options opened

  • Marriage migrants:
    • Simplified naturalization (1998, now by application and review)
    • Simplified naturalization after the marriage ends or while raising a child (2004)
    • Spousal permanent residence (2005)
    • Stay for parents raising a child (2007, F-2)
    • Stay after a marriage ends through no fault of one's own (2011.12, new F-6; its other items reclassified from F-2)
  • Co-ethnic Koreans and others: co-ethnic permanent residence (2008, F-5-6 and F-5-7; ethnic Chinese long-term residents F-5-8)
  • Dual citizenship (2011.1, a pledge not to exercise the foreign nationality in Korea): covers naturalizing marriage migrants still in the marriage, people of special merit or talent, and adoptees and people aged 65+ returning to reinstate Korean nationality, among others. General naturalizers were excluded.

Where people ended up

  • Marriage migrants overall: stayed on visas and then naturalized; few moved to permanent residence (5–10% around year 8).
  • Vietnam: the highest naturalization; filled the drop in Korean-Chinese naturalizations and kept the total steady.
  • Korean-Chinese:
    • As a co-ethnic population, many received permanent residence through the 2008 co-ethnic route.
    • Among marriage migrants, naturalization dominates; among those not naturalized, about one-third hold permanent residence.
  • China (non-Korean): loses original citizenship on naturalization, and has the highest PR share among marriage migrants (17–26%).
  • Philippines: wants naturalization about as much as Vietnam, but both naturalization and PR are low; about 70% stay on visas. The reason is still unknown.
  • Dual citizenship: marriage migrants were eligible, but no increase in their naturalization was found. The clear effect appears in nationality reinstatement.

Naturalization

  • Requirements tightened over this period: a written-test reform in 2009, the test exemption limited to spouses still in the marriage from 2010.7, and stricter marriage-visa screening in 2014.
  • Women who entered from 2009 on show lower naturalization at the same length of stay, but they catch up over time, so naturalization came later rather than less often.

Key reforms (checked against statutory texts)

In forceReformLegal basisOption opened
1998.6.14Automatic nationality for foreign wives abolished; simplified naturalization after two years of residence in marriageNationality Act, Law 5431Naturalization by application and review
2002.4.18Permanent residence (F-5) created; spouses of citizens hold residence (F-2), PR after five years on F-2Immigration Control Act Enforcement Decree 17579PR (five-year route)
2004.1.20Simplified naturalization extended to those whose spouse died or disappeared, whose marriage ended through no fault of their own, or who raise a childNationality Act, Law 7075Naturalization after the marriage ends
2005.9.25PR route for spouses of citizens and PR holders after two years of stayDecree 18934Spousal PR (F-5-2)
2007.6.1Parents raising a child born in marriage to a citizen added to F-2Decree 20076Stay after divorce while raising a child
2008.7.3Co-ethnic PR routes created (two years on F-4; co-ethnics meeting naturalization requirements; ethnic Chinese)Decree 20892Korean-Chinese PR, unrelated to marriage
2011.1.1Dual citizenship by pledge (naturalizing marriage migrants in the marriage, talent, adoptees and 65+ reinstatements, among others)Nationality Act, Law 10275Keep original nationality after naturalization
2011.12.15Marriage-migrant visa (F-6) created; no-fault marriage breakdown added, other items reclassified from F-2Decree 23274Stay after marriage breakdown without children

Other changes in the same period: naturalization written-test reforms (2009.1.1 exemption for integration-program graduates, 2009.6.1 multiple-choice format, 2010.7.1 exemption limited to spouses in the marriage), marriage-visa screening (introduced 2011.3.7, income and Korean requirements 2014.4.1), PR requirements put into regulation (2018.9.21), and PR required before general naturalization (2018.12.20).

Key figures

Reinstatement: US cases jumped about 15-fold in 2011; Japan did not move

Annual nationality reinstatements. Separate y-scales. Dashed = 2006–10 mean; dotted = 2011 dual-citizenship reform

Data: Ministry of Justice Immigration Statistics Yearbook (2006–10), KOSIS naturalization and reinstatement by nationality (DT_1B040A19, 2011–25). 2008–09 shown as a 2-year annual mean. Result file: results/revalidation_2026/reinstatement_did_20261109/R0_panel.csv (analysis 69).

Naturalization: Vietnam filled the Korean-Chinese decline; total stays above 10,000

Mean annual naturalizations by origin. Excludes reinstatement and the 2009–10 surge

Source: 2007 Statistical Yearbook cumulative table (2006–07), KOSIS DT_1B040A19 (2011–25). 2016–20 China and Korean-Chinese exclude 2017 (KOSIS duplicate estimate). The 2006–07 origin table sums above the published total, so the Korean-Chinese decline may look slightly larger than it is.

About 8 years after entry, 40–67% had naturalized; PR stayed at 5–10%

Status of marriage-migrant women at the survey (weighted %)

Data: National Survey of Multicultural Families 2015, 2018, 2021, 2024 microdata (women, nine origins, household weights). Status is naturalization and visa at the survey. Synthetic cohorts: each survey compares the same entry-year group, not the same people. Result file: results/revalidation_2026/options_to_status_20261109/A1_cohort_status_by_wave.csv (analysis 71).

Why later entry cohorts naturalize less: time on the marriage visa, not a move to PR

Cumulative naturalization by entry cohort (Figure 2 of report v2; choose the survey)

At the same years since entry, the 2009–10 and 2011–14 cohorts naturalize much less for the Philippines, mainland China and Korean Chinese, and almost the same for Vietnam. Data: 2015, 2018 and 2021 survey microdata (all respondents born in each origin, both sexes, household weights, reconstructed from entry and naturalization years; 2021 shown first). Open circles: fewer than 20 cumulative events or 30 respondents. Result file: results/revalidation_2026/entry_cohort_cumulative_20261026/cumulative_curves_4origins.csv. Source: report v2 Figure 2, code/revalidation_2026/57_entry_cohort_cumulative_4origins.py.

Status and plans of the same cohorts at each survey (%; choose the survey)

Status is measured at each survey, so later cohorts have had less time in Korea (see mean years). Figures quoted above refer to the 2021 survey. Plans are asked of the non-naturalized only; in 2015, "no plan" combines "PR only" and "neither". * fewer than 30 respondents answering plans. The 2024 survey has no naturalization year, so it appears in the table but not the figure. Data: 2015, 2018, 2021 and 2024 survey microdata (both sexes, household weights). Result file: results/revalidation_2026/cohort_status_all_waves_20261110/S1_cohort_status_visa_plans_by_survey.csv (script 78; 2021 matches Table F2 of report v3).

Status about eight years after entry, by origin (naturalized / PR / visa, %)

OriginEntered 2005–08 (2015 survey)2009–11 (2018)2012–14 (2021)2015–17 (2024)
Vietnam76/1/2356/2/4257/2/4067/2/31
Korean-Chinese77/6/1849/12/3947/14/4047/6/47
China (non-Korean)60/12/2832/26/4220/10/7026/15/59
Philippines53/3/4428/4/6925/2/7325/6/69

PR share among non-naturalized 2009–11 entrants: Korean-Chinese 40% (2021) and 31% (2024); China 33→30%, Philippines 8→9%, Vietnam 7→0%.

Annual PR and naturalization rates (2015 survey, years 2–9 of residence, %, PR / naturalization)

Origin1995–20042006–082009–102011–15
China (non-Korean)0 / 12.50.5 / 15.02.3 / 16.33.7 / 8.7
Korean-Chinese0 / 11.40 / 16.20.7 / 23.12.0 / 15.1
Vietnam0.6 / 10.00.2 / 9.10.1 / 14.80.3 / 13.4
Philippines0 / 15.60.2 / 12.80.4 / 14.70.4 / 5.8
Other (Japan, Taiwan, etc.)0.7 / 1.34.0 / 2.83.2 / 5.22.2 / 5.5

Status after the marriage ended (2018 and 2021 surveys, weighted %)

Marriage endedRespondentsNaturalized beforeNaturalized afterPRVisaNaturalized within 3 yrs*
2004–0614069260.5534
2007–1041971200.6850
2011–145778082.5926
2015–211,1347621.72014

* Of those not yet naturalized when the marriage ended. Recent dissolutions have had little time (2.5 years on average for 2015–21), which raises the visa share. Divorced or separated: 79/9/1.5/11; widowed: 62/17/1.7/19 (naturalized before / after / PR / visa).

Interpretation and limits

Files: docs/개혁연혁-법령원문-확인-20261109.md, docs/법령원문_20261109/, code/revalidation_2026/ (scripts 69–76), results/revalidation_2026/.