Migrant workers in East Asia

I · AT A GLANCE

How the law applies to each group

The unit of comparison is the worker group, not the state, because each jurisdiction runs separate legal regimes for different migrants. The sharpest contrast lies within Taiwan: fishers hired in Taiwan fall under the Labor Standards Act, while crew recruited abroad for the distant-water fleet are governed by fisheries regulations with their own, lower wage floor.

II · COMPARE

Build your own comparison

Choose which features matter for your question by moving the sliders. A weighted comparison appears once you set a weight. Each feature counts 2 when the protection is full, 1 when partial and 0 when absent, following the codebook in section XI.

III · INTERNATIONAL LAYER

Treaties, and what stands in for them

Taiwan is not a member of the United Nations or the International Labour Organization and cannot ratify their treaties. It has instead written the two UN covenants into domestic law, and its fishing labor rules are increasingly shaped by importing states: the United States lists Taiwan-caught fish as produced with forced labor, and the European Union's yellow card of 2015 to 2019 and its new Forced Labour Regulation reach the fleet through trade. Hong Kong is bound through extension by the United Kingdom before 1997 and by China since.

IV · JURISDICTION PROFILES

Each jurisdiction in detail

V · PEOPLE AND JOBS

Who the workers are

Headline counts for each jurisdiction. Official figures on fishers are sparse and dated, and estimates by advocacy organizations for the distant-water fleets are higher; each figure shows its year and source.

VI · FRAMEWORKS AND AGREEMENTS

The normative framework, by level and nature

Instruments that bind or guide the four jurisdictions on these workers, classified by level (international, regional fisheries bodies, measures by importing states, bilateral) and by nature. Importing-state measures are listed as a level of their own because, for Taiwan's fleet, they do much of the work that treaties do elsewhere. Click a cell in the map to filter the list.

A memorandum of understanding usually creates no binding legal obligation; a treaty or agreement does once in force. A measure of a regional fisheries management organization binds its members and cooperating parties, including Taiwan as a fishing entity.

VII · DOMESTIC LAWS

Domestic laws, with links to the official texts

The statutes, regulations and standard contracts that govern each worker group. Links lead to the official text where a stable address exists, and otherwise to the official legislation database, where the title or original-language title can be searched. Official English translations are for reference; the original-language text prevails.

VIII · CORRIDORS

Where the workers come from

The largest origin countries in each jurisdiction, from the best official figure we could obtain. Indonesia, the Philippines and Vietnam appear in every corridor.

IX · EXPLORE THE DATA

Explore patterns across corridors

Each row is one corridor: a country of origin and the jurisdiction its workers go to. Ask whether corridors with bilateral instruments carry more workers, or whether origins that ratified the UN migrant workers convention are more often covered by one. Results show confidence intervals, p-values and a plain reading.

With few corridors and different populations counted in each jurisdiction, these are descriptive associations, not tests of cause. Large corridors may attract agreements rather than the other way round.

X · LEGAL DEVELOPMENTS

What changed, and when

XI · METHOD AND CITATION

How the data was built

The unit of analysis is the legal regime that applies to a defined group of migrant workers, on paper (de jure), as of the date shown. Enforcement (de facto) is not coded; where reports show a gap between law and practice, the note to the cell says so. Eight features concern labor protection; two (a route to long-term residence and family accompaniment) concern immigration status, following the categories of the International Organization for Migration's International Comparative Study of Migration Legislation and Practice (2002). Each cell is coded by the written rule below and records its note, its source, the date coded and whether it rests on the primary legal text (P) or on a secondary source or a statutory citation not yet checked against the official text (S).

Limitations: official English translations are unofficial in law and the original text prevails; regulations and ministerial notices change faster than statutes; official statistics on fishers are sparse and dated, and the populations counted differ between jurisdictions. Treaty status should be confirmed in NORMLEX and the UN Treaty Collection before citation.

Reliability and replication

• Coding protocol: one coder applied the rules above to 70 cells (7 groups × 10 features). On 31 October 2026 a second reader checked 30 cells against the official legal texts in the original language, quoting the provision; those cells are marked P ("partly" where part of the note still rests on a secondary source or on our reading), and one cell was recoded. A full independent second coding is planned, with agreement reported as Cohen's kappa. Cells marked S should be treated as provisional.
• Feature selection: the eight labor features follow the core ILO standards for fishing (C188) and domestic work (C189). The mobility, residence and family features follow the restrictions that define temporary labor migration programs in Asia, namely a visa tied to one employer, no change of sector, no route to long-term residence and no family reunion (Oberoi and Sheill 2024; OHCHR 2022), and the immigration-law categories of IOM (2002).
• Corridor analyses treat each origin–destination pair as a case. Pairs within one jurisdiction are not independent and each jurisdiction counts a different population, so the presets include the destination as a control; the p-values do not correct for clustering and should be read as descriptive.
• Many tests can be run on the same data. Treat any single significant result as a lead to investigate, not a finding; the statistics engine applies Holm's correction to pairwise comparisons.
• Any weighted comparison depends on the weights chosen. The stability check under the comparison shows how often each group scores highest across 10,000 random weightings.
• Replication files: the codebook, every coded cell with its note and source, treaty status, instruments, domestic laws, corridors and developments can be downloaded below as CSV and JSON. Each release carries a version number and a dated change log.

How to cite

XII · FROM RESEARCH TO PRACTICE

From research to practice

Each jurisdiction profile above lists the arguments the law opens and the avenues for using them, with a printable one-page brief. Our overview of legal mechanisms in East Asia explains labor dispute systems, human rights bodies, home-country channels and international and trade routes.

XIII · SOURCES AND DOWNLOADS

Sources and downloads

Static CSV files: Coded cells with notes and sources · Codebook · Treaty status · Instruments · Domestic laws · Corridors · Legal developments. Licensed CC BY 4.0.

This page summarizes laws for research and advocacy. It is not legal advice. Found a mistake? Write to secretariat@theimr.org.