Little ghosts

ผีน้อย, usually rendered as little ghosts, is a Thai term used for undocumented Thai migrants abroad, particularly in accounts of South Korea.

It is a nickname, not a category that tells us somebody’s job, route into the country or legal rights. Nor does it establish that everyone described by it uses the name themselves.

The distinction matters because our earlier version made the word do too much. It turned a population figure into a workforce, tourist arrival into everybody’s route, and precarious status into an absence of every protection. The public record supports a more specific account.

Count residents before counting workers

South Korea’s Ministry of Justice recorded 115,851 Thai nationals staying without lawful status at the end of 2025, out of 357,598 foreign nationals in that category. Thailand was the largest nationality group, accounting for 32.4%.

Those are residence statistics. They do not establish that every person counted was employed, what they earned or which job they held. A lawful residence category also does not automatically tell us whether a particular job was authorised.

The same original yearbook gives this Thai series:

The count fell in 2024 and 2025. It had risen in 2022 and 2023, so the previous claim of a decline every year was wrong. These are successive stocks of people present at a date; their difference is not a count of voluntary returns or proof that one policy caused the change.

There are newer aggregate statistics. The September 2026 monthly report records 344,387 foreign nationals without lawful stay across all nationalities. That is a different period and population from the Thai year-end series above. It cannot update the Thai count by itself, and we do not divide figures from different dates to manufacture a current Thai share.

A visit and a job require different permissions

The official K-ETA eligibility page lists Thailand with a 90-day permitted stay and describes short visits for travel, relatives, events, conferences or business without commercial activity.

That does not authorise taking a job. A visa waiver, travel authorisation, admission to the country and permission for particular work are separate questions. The appropriate official entry and work requirements need checking before travel; a broker’s assurance cannot substitute for them.

Overstaying a short visit is one possible route into irregular residence. It is not the only possible route. The Korean yearbook records people without lawful stay under categories including E-9 employment, study and training as well as B-1 and B-2 short visits. Those national totals do not identify every Thai person’s history.

The earlier assertion that the entire route involved no agent, deception, smuggling or trafficking is withdrawn. A lawful border crossing cannot tell us whether recruitment was coercive or deceptive. An immigration category is not evidence that either occurred, or that it did not.

The legal work routes are different

E-9 is the non-professional employment route used under the general Employment Permit System, or EPS. Korea’s government legal-information service describes a sequence involving a Korean-language test, jobseeker registration, a contract, visa issuance and employment training. A test result is one stage of that process, not a job offer.

E-8 concerns seasonal work. The Thai official account discussed below describes agricultural and fisheries placements arranged through agreements with Korean local authorities. It is a separate route from EPS.

E-7-4 concerns skilled-worker status. An official Korean notice documents an expansion in September 2023; it was not created in 2026. Thailand’s account of bilateral talks in May 2026 also mentions support for E-9-to-E-7-4 changes.

That establishes an existing policy route, not automatic eligibility for every worker. Current criteria and the applicant’s record still matter. Neither the route’s existence nor a falling residence count demonstrates that it deterred people from leaving employers.

The old comparison of a 10,000-place allocation with demand in the hundreds of thousands is removed. It did not establish a matched application cohort, programme, period, waiting time or number of available jobs. We cannot infer an individual’s alternatives from that comparison.

What the wage figure can tell a family

Korea’s Minimum Wage Council lists the 2026 hourly minimum as KRW 10,320. Its published monthly equivalent is KRW 2,156,880, calculated on 209 hours.

That reference amount is not a promise of take-home pay, a measure of what undocumented workers receive or an estimate of money available to send home. Contracted hours, actual payment, deductions, accommodation, living costs and recruitment expenses need their own evidence.

A useful comparison starts with a dated contract and costs for the same period, rather than converting an advertised monthly amount into baht and treating the difference as profit. This is an editorial checklist, not a finding about every worker’s decision.

For the same reason, a residence count cannot be multiplied by an assumed remittance to produce a Korea-to-Thailand total. Article 37 explains the limits of its separate household-transfer evidence.

Status does not erase every protection

Seoul’s public industrial-insurance guide expressly includes workers without the proper visa among those eligible to apply for industrial-injury compensation. It also describes coverage exclusions. The workplace, employment relationship, injury and applicable rules still need assessment.

A Korean labour ministry explanation dated 12 October 2022 similarly states that industrial-accident treatment and compensation do not distinguish workers by nationality or lawful/irregular status. The same explanation says the compensation agency supplies information when requested under immigration-law provisions.

These sources refute a blanket claim of no compensation route. They do not provide a guarantee about an individual’s award, confidentiality or immigration consequences. The 2022 explanation is a dated official position, not a comprehensive statement of every current procedure.

Having a legal route and being able to use it are also different. Language support, documents, representation and practical access need investigation. Article 89 cannot answer those questions through a general theory about who gets protection.

Our old comparison between foreign workers’ workforce share and fatality share did not establish an undocumented Thai worker’s separate risk, let alone its cause. We have removed the unmatched ratio and an inadequately sourced death anecdote. The unverified annual total of Thai deaths worldwide is also removed; this review did not establish either that total or the claim that no breakdown existed.

What the four-province restriction covered

A 13 May 2026 Thai government statement, checked with the Department of Employment, describes suspension of seasonal agricultural and fisheries recruitment under E-8 from Udon Thani, Khon Kaen, Chaiyaphum and Maha Sarakham, for 1 January–31 December 2026.

It attributes the measure to some participants leaving their employers. This is the Thai government’s account; the original Korean decision and the individual local agreements were not obtained for this review.

The restriction concerned that seasonal channel. It does not establish a general ban on Thai employment, the closure of EPS, or the guilt of every prospective applicant from those provinces. We have removed the unsupported numerical absconding threshold and the verdict that one entire group was paying for its neighbours’ choices.

Thailand’s account of talks on 18 May 2026 lists the restriction, voluntary return, human rights, workplace safety and visa transitions among the subjects discussed. Discussion is not proof of a subsequent concession or measured improvement.

Anyone considering a placement needs the current Department of Employment notice and the rules of the relevant programme. The May announcement should not be treated as a promise about recruitment after its stated year ends.

Common misconceptions

“The Thai residence figure counts illegal jobs.” It counts people in a residence category at a specified date. Employment and permission for a particular job require separate evidence.

“A tourist entry can become a work permit simply by staying.” Short-visit permission does not confer employment permission. EPS has a separate documented process.

“An undocumented worker has no protection of any kind.” Official compensation guidance contradicts that blanket claim. Coverage, access and immigration consequences still need individual assessment.

“The 2026 restriction stopped every legal route from Thailand.” The cited Thai announcement concerns E-8 seasonal recruitment from four provinces, rather than all Thai workers or every work visa.

The useful questions are concrete: which people were counted, on what date, under which permission, with which protection, and with what evidence of its use. The nickname answers none of them on its own.