When the market stopped watching

In April 2015, the European Commission gave Thailand a yellow card — the strongest formal warning it issues over illegal, unreported and unregulated fishing that carries no penalty attached.

On 8 January 2019, it was lifted.

What happened in between is the single strongest piece of evidence I have found against my own argument in article 76 Nothing lasts long enough to build, and what has happened since is the strongest piece of evidence for it.

What Thailand actually did

I want to give this its full weight, because it is genuinely impressive and it is not how this story usually gets told.

Labour registration in Thai fishing went from almost non-existent in 2014 to systematic registration and regular inspections by the end of 2019. That is the ILO’s Ship to Shore Rights Project assessment, not a government claim.

Thailand ratified ILO Convention C188 on work in fishing, immediately before the card was lifted.

Labour standards were written into the fisheries and migration statutes themselves, with penalties for vessel and factory owners violating labour protections and for employing migrant workers without valid permits.

And migrant fishing workers were required to be paid into bank accounts by electronic transfer — which sounds administrative and is one of the most effective anti-exploitation measures available, because a wage that moves through a bank leaves a record and a wage handed over on a dock does not.

Five years, from near-nothing to a functioning inspection regime. Article 76 argued that Thailand cannot build institutions because nothing lasts long enough to build them. This is a country building an institution quickly and well.

And then

In February 2024, the Thai Parliament accepted the principle of eight amendment drafts and sent them to committee. Not a narrow vote, not a partisan one. Bipartisan.

On 25 December 2024 the House of Representatives passed the bill at second and third reading and sent it to the Senate — reported at 239 votes in favour.

And then the second chamber refused the central provision. On 25 February 2025 the Senate resolved to amend the bill, striking out the exception that would have allowed surrounding nets of mesh smaller than 2.5 centimetres to be used at night. That exception is the heart of it. The section as it stands bans the practice outright; the amendment would have permitted it subject to conditions set by the Minister.

Because the two chambers disagreed, the bill went to a joint committee of both houses — on 19 March 2025 the House declined the Senate’s version by 398 votes to none, which under the constitution sends a bill to a committee of eleven members from each chamber. That committee first met on 17 July 2025 and finished on 3 September 2025. Its text restored the night-net exception in a narrower form, outside twelve nautical miles and conditional on research, and the Senate approved it on 16 September 2025 and the House on 17 September 2025 — the dates the Department of Fisheries gave in its own statement of 30 September 2025, which described the bill as with the Secretariat of the Cabinet for the steps that lead to publication in the Royal Gazette.

A year later it is not in force. The Council of State’s consolidated index of legislation, which carries Acts published as recently as 28 August 2026, lists the 2015 Decree with no amending Act against it. Where the bill sits between parliament and the Gazette is not something this site can establish from a primary source, and it does not guess. In February and March 2026 the National Fisheries Committee took up, and then postponed, the research guidelines on which the restored exception depends.

What the amendment does on labour can now be stated from the bill itself rather than from anyone’s characterisation of it. The House-passed text, 71 sections long, is set out clause by clause in the Senate Secretariat’s briefing document, alongside the current law it replaces. Three things in it are the labour story:

• It repeals sections 10/1, 11 and 11/1 of the Decree — the provisions that made seafood-processing factories notify the fisheries authority before operating, that let an administrative committee order a plant to stop for ten to twenty days for employing foreign workers without permits, escalating to closure and a ten-year bar on repeat, and that applied the same stop-orders to plants committing the child-labour offences in sections 44 and 49 to 51 of the Labour Protection Act. The bill’s own explanatory note gives the reason: these are now covered elsewhere — by the Labour Protection in Fishing Work Act B.E. 2562 (2019), the 2022 ministerial regulation on sea-fishing labour, the Labour Protection Act as amended in 2023 and the Foreign Workers Management Decree. That is the government’s case, stated fairly. What it removes is the fisheries regulator’s own power to shut a plant; what remains is the labour ministry’s ordinary enforcement.
• It amends section 82 so that a registered Thai fishing vessel whose crew numbers officials can check under the navigation law no longer has to present its crew list and seafarers’ books before leaving port. The pre-departure crew check was one of the measures that made the 2015 settlement inspectable at the quayside; the amendment does not abolish the records, it removes the moment at which they had to be shown.
• It restores the small-mesh night-net exception in section 69, in the joint committee’s narrowed form.

So the honest statement is neither that the rules have been unwound nor that the attempt has failed. Parliament has finished: both chambers have approved a text that takes the fisheries regulator out of the seafood-processing plant and out of the pre-departure crew check. It is not yet law, and nothing on the water has changed that this site can verify.

The industry’s case, taken seriously

There is a real argument on the other side and it deserves stating properly, because the version of this article that treats every amendment supporter as a villain is both wrong and useless.

The 2015 Fisheries Decree was written fast, under external pressure, to satisfy a foreign regulator. Legislation drafted in those conditions is rarely well-calibrated, and the complaint that the resulting rules are overly restrictive and complex, and damaging to the profitability of Thai fleets, is not obviously false.

Small-scale Thai fishers have genuine grievances about compliance costs and administrative burden that fall hardest on the operators least able to absorb them — which is, incidentally, article 89’s pattern showing up inside the reform rather than outside it.

A rule can be both necessary and badly drafted. Fixing the drafting and removing the protection are different projects, and the argument is about which one is happening.

What it costs if it goes wrong

Analysis has put US$3.3 billion of seafood trade at risk — exports to six countries plus the EU bloc, representing just under 60% of Thailand’s seafood trade on 2022 figures.

That is the practical answer to anyone who thinks this is a human rights story that businesses can ignore. The buyers are the enforcement mechanism, and they have demonstrated once already that they will act.

Human Rights Watch has publicly urged the EU and Thailand to address migrant worker rights within their trade negotiations, which tells you the issue is live in exactly the forum where it has historically had teeth.

The finding, which is uncomfortable

Here is what I think this case actually establishes, and it is more precise and more useful than either “Thailand can’t reform” or “Thailand did reform.”

Thailand reformed when a market required it, and began trying to reverse when the market stopped watching.

Not because of any failure of capability. The capability was demonstrated — comprehensively, in five years, on a hard problem.

What was missing was a domestic constituency that outlasted the external pressure.

And here is the part that connects it to everything else on this site.

The workers these rules protect cannot form unions.

That is a documented gap in the post-2019 settlement — migrant workers in Thai fishing lack the ability to organise. Article 84’s delivery riders at least got a hundred people outside a ministry. Article 96’s auto workers have factories, contracts and a severance ladder.

Migrant fishers have no vote, no union, and largely no documentation — with 1.8 million undocumented Myanmar nationals in Thailand alone, across fisheries, construction and elsewhere.

So the only constituency that has ever effectively spoken for them was a European customs classification.

Article 89 The wrong way round argued that Thai protection varies inversely with exposure because categories get drawn around people who already had options. This is the extreme case: a workforce with no domestic political existence at all, whose protections were granted by a foreign trade instrument and are being contested now that the instrument is quiet.

That is not a Thai failing specifically. It is what happens to any workforce that cannot vote, cannot organise and cannot be counted. Thailand is simply where it is currently visible.

What would make it stick

Let them organise. The single change that would make every other protection self-sustaining, because a right somebody can defend does not depend on who is watching. It is also the change with the least chance of happening.

Keep the electronic wage requirement whatever else changes. It is cheap, it is administrative, it leaves a record, and it is the most effective single measure in the whole package.

And buyers should say what they require, publicly, now. The enforcement in this system has always come from purchasers. Waiting for another card is a slower and more damaging mechanism than telling suppliers in advance what will and won’t be bought.

Common misconceptions

“Thailand never fixed its fishing industry.” Labour registration went from near-zero in 2014 to systematic inspection by 2019, and the EU lifted its warning in January 2019.

“The reforms were cosmetic.” The ILO’s own project assessment recorded substantial improvement. Gaps remained, notably on organising rights and abuse reporting.

“The amendments are a partisan move.” The principle was accepted without a partisan split, the House passed the bill in December 2024, the Senate struck out its central provision in February 2025, and both chambers approved the joint committee’s compromise in September 2025. The disagreement has run between the chambers, not between parties.

“This is only a human rights issue.” Just under 60% of Thai seafood trade, around US$3.3 billion, goes to markets that have acted on this before.

“Everyone backing amendment is acting in bad faith.” The 2015 Decree was drafted quickly under foreign pressure and complaints about complexity and cost are not obviously unfounded. The question is whether fixing drafting and removing protection are being distinguished.