EU-Mercosur: The standards gap

In the EU, 38% of laying hens are still kept in cages. In Argentina, the figure is 89%. In Uruguay, 88%. In Brazil, 95%.

To qualify for duty-free access to the EU market, Mercosur egg producers must comply with EU laying-hen welfare standards. The first time the EU has conditioned tariff access on animal welfare. A real precedent.

So what is the problem, one might ask?

The problem is that it only applies to shelled eggs. It does not apply to egg products. And guess what? The EU barely imports any shelled eggs from Mercosur.

I want to start here because this small example contains in miniature the whole shape of the standards conversation around the EU-Mercosur deal — and the reason that conversation has been so frustrating to follow.

The distinction the conversation usually skips

There is a distinction in trade law between product standards and process standards.

Product standards are about what a thing physically is when it crosses the border. These are properties of the product, testable at the point of import.

Process standards are about how the thing was made. Were the hens kept in cages? Were the cattle treated with hormones banned in the EU? Were the crop fields sprayed with pesticides Europe banned twenty years ago? These are properties of the production system and they cannot, in most cases, be tested at the border. They have to be certified by the exporting country's regulatory system, and that certification has to be trusted.

When the European Commission says that "EU sanitary and phytosanitary standards are non-negotiable" — this is true.

Yet, the question critics are mostly asking is whether EU process standards — the ones European farmers actually have to comply with on their own farms — apply to imported products in any meaningful way. The answer to that question is more complicated, and the Mercosur deal does not simplify it.

Take the Commission's case seriously

Before going further, the Commission's framing deserves to be stated as fairly as it can be. Imports of food and agricultural products into the EU are subject to substantial regulation. There are border controls. There are audits of exporting countries' regulatory systems. There are maximum residue limits for pesticides. There are bans on specific substances — the EU began restricting growth-promoting hormones in 1981, fully implemented an import ban in 1989, and permanently banned estradiol-17β — the substance at the centre of the recent Brazilian audit findings — in 2003. There are EU-led sustainability dialogues with major trading partners. The agreement with Mercosur reaffirms all of this, adds the laying-hen welfare condition for shelled eggs, includes a "first-of-its-kind joint AMR cooperation commitment," and establishes structured dialogues on animal welfare, biotechnology, food safety and antimicrobial resistance.

That is a real framework. It is not nothing.

What follows is about what that framework does not cover.

1. Substances

This is where the data is sharpest, because pesticides and hormones are the most measurable category.

As of late 2024, Brazil had authorised 330 pesticide active substances. Sixty percent of them were not approved in the EU. Of the 421 substances approved in the EU, 73% were not authorised in Brazil. So the regulatory regimes have diverged significantly: each side bans many substances the other allows, and Europe's bans are systematically tighter on substances classified as carcinogenic, mutagenic, toxic to reproduction, or hazardous to bees and pollinators.

Some of the EU-banned substances most widely used in Brazil are atrazine (banned in the EU since 2004 over groundwater contamination concerns), acephate, chlorothalonil (classified as a probable human carcinogen by EU authorities), and chlorpyrifos (banned over developmental neurotoxicity and DNA damage concerns). In 2020 alone, 33,300 tonnes of atrazine, 29,900 tonnes of acephate, 24,100 tonnes of chlorothalonil and 8,800 tonnes of chlorpyrifos were sold in Brazil — much of it manufactured by EU-headquartered companies for export.

The EU has a tolerance system for residues of these substances in imported food. Where a substance is banned for use in the EU but residues are detected in imports, the Commission sets "import tolerances" — maximum allowed residue levels — based on dietary exposure assessments. This system means EU consumers can be exposed, through imported food, to residues of substances Europe has banned for its own farmers.

The Veblen Institute, foodwatch and PAN Europe commissioned an independent legal opinion on this practice. The opinion's central conclusion is uncomfortable: for pesticides banned on public health grounds, the Commission's practice of setting import tolerances may not just be problematic but illegal under existing EU law. The MRL Regulation, the legal opinion argues, requires the automatic deletion of residue limits when an active substance is revoked. The Commission has not been doing this. The argument elevates the standards conversation: critics are not asking for new legislation, they are arguing that current EU practice already fails to apply current EU law.

In January 2025, the Commission proposed banning import residues of three specific fungicides (carbendazim, benomyl, thiophanate-methyl). Two of the three are already banned in Brazil. The Veblen analysis estimates the proposal covers about 22% of EU-banned substances.

The hormones case is sharper still, because here we have direct enforcement evidence.

The European Commission's DG SANTE conducted an audit of Brazilian beef export controls in May-June 2024. The audit found that Brazil's regulatory system "could not reliably attest to operator compliance" with the EU's ban on oestradiol 17β — a hormone banned in the EU since 2003 on cancer-risk grounds, restricted in EU livestock since 1981. Brazil pledged to suspend exports of female cattle (males are not treated) and put a new traceability protocol in place by January 2025.

In October 2025, a Commission audit team returned to inspect the new protocol. They visited a sample of farms enrolled in it. On at least one farm, MAPA's own controls had identified that 179 female cattle treated with oestradiol during autumn 2024 had been wrongly enrolled in the export protocol. For 174 of those cattle, the certifying entity had nonetheless issued transaction certificates making the animals eligible for slaughter for the EU market in May and July 2025. Across the audited farms, the team also found that prescriptions for oestradiol-containing veterinary products — required by the new protocol — could not be produced, and that this had not been flagged as a deficiency by either the certifying entity or by MAPA's own controls.

So the chronology is: 2024 audit identifies the problem, Brazil pledges fix, January 2025 protocol established, October 2025 audit finds the protocol is failing in essentially the same pattern as the system it replaced.

In November 2025, the EU's Rapid Alert System for Food and Feed (RASFF) issued a recall notice for frozen Brazilian beef containing oestradiol 17β. The contaminated consignments had entered the EU via Rotterdam and reached twelve member states. In Italy, an importer cooked and served the affected beef before the recall reached him. The Commission characterised the volumes as small — about 0.1% of Brazilian beef imports during the relevant period — and said the consignments did not pose immediate health risks. Both points are accurate. They are also beside the point. A control system that allows banned-substance contamination to reach consumers is a control system that has failed at its core function, regardless of whether the failure rate is one in a thousand or one in ten.

This is the most concrete evidence available that the Commission's "EU SPS standards remain fully applicable" framing answers a different question from the one being asked. The standards remain on paper. Their application to imports, in this specific case, has documentably failed repeatedly in less than two years.

2. Process

The animal welfare case is where the product/process distinction is cleanest.

EU welfare law covers slaughter conditions for imports — animals must be killed in conditions equivalent to EU slaughterhouse standards. It does not, with the laying-hen exception noted in the opening, cover the animal's life before slaughter.

The practical consequence: a Brazilian sow can spend her entire reproductive life confined in a gestation crate — a practice illegal in the EU since 2013 — and her meat can still enter the EU legally, provided she was killed in conditions equivalent to EU slaughterhouse standards. A Brazilian broiler chicken can be raised at stocking densities prohibited in Europe, with mutilations performed without pain relief, and the meat enters legally. The hens that produce egg products (as distinct from shelled eggs) can be kept in conventional battery cages — also illegal in the EU since 2012 — and the products enter legally.

The European poultry processors' association AVEC, hardly a fringe NGO voice, put it bluntly in February 2025: "Audit reports performed by the EU Commission in 2013 and 2017 show that Brazil does not respect EU rules. The EU poultry meat sector feels betrayed by the Commission. Our sector has been sacrificed to satisfy the interests of bigger players." That is an industry quote, not an environmentalist one — and worth noting because it complicates the "farmers versus environmentalists" frame that the standards conversation tends to default to.

Environmental compliance on production — nitrate management, manure handling, biodiversity protections, air quality — is a parallel issue: EU farmers carry significant regulatory cost burdens that Mercosur producers do not face equivalently. The data here is patchier than for substances or animal welfare, and the topic deserves its own treatment rather than a paragraph at the end of this section. I will return to it.

3. Enforcement

The Commission's response to most standards concerns is, in effect: the rules apply, we audit, we enforce. The hormones case shows what that enforcement looks like in practice when the underlying traceability systems are weak.

Brazil has approximately 1,220 cattle farms approved to deliver to 53 EU-approved slaughterhouses. The 2024 DG SANTE audit found that the system in place to verify which animals on those farms had been treated with banned substances was inadequate. The 2025 follow-up audit found that the replacement system was inadequate in the same way. The pattern is not a one-off compliance failure; it is a structural feature of attempting to enforce EU process standards through a third-country regulatory system that has different priorities, different resources, and different incentives.

The deal's response to this enforcement question is a structured dialogue on animal welfare, a "first-of-its-kind joint AMR cooperation commitment," and a Trade and Sustainable Development chapter with — per ClientEarth's legal analysis — a dispute mechanism that "cannot impose sanctions for non-compliance." A separate bilateral safeguard regulation, agreed between Council and Parliament in December 2025 and formally adopted in early 2026, allows the EU to suspend tariff preferences temporarily on sensitive agricultural imports from Mercosur. Its operative quantitative triggers are volume-based and price-based: as a rule, a 5% increase in preferential import volumes on a three-year average, or a 5% price undercut, is treated as sufficient grounds to launch an investigation. Parliament added a reciprocity provision under which the Commission "shall initiate an investigation and adopt safeguard measures where there is credible evidence that imports benefiting from tariff preferences do not meet equivalent environmental, animal welfare, health, food safety, or labour protection requirements." Whether that provision becomes operative in practice depends on how "credible evidence" is interpreted, and on whether the Commission chooses to initiate investigations on standards grounds rather than waiting for member states or industry to do so. The 2024 and 2025 Brazilian audit findings were credible evidence by any reasonable standard. No safeguard investigation followed.

4. Architecture: mirror clauses and the future

The technical name for the policy tool that would close most of the gaps described above is a "mirror clause" — a provision requiring imports to comply with the same production standards as domestic products, not just the same product standards. The EU has implemented mirror-style measures in narrow areas before: restrictions on imports of hormone-treated animal products (Directive 96/22/EC) and on products from animals not slaughtered to EU welfare standards (Regulation EC 1099/2009). The legal precedent exists. The political will to extend it has been intermittent.

France has made mirror measures a central condition for its support of the Mercosur deal. The French government, joined at various points by Poland, Italy, Austria, Belgium, Luxembourg and Slovakia, has called for explicit mirror clauses on pesticides, animal welfare, and environmental standards. The Commission's "Vision for Agriculture and Food," published in February 2025, commits to a 2025 assessment of a principle prohibiting entry of products produced with hazardous EU-banned pesticides — a step toward mirror measures, in a limited form, on a parallel timeline.

The Mercosur deal does not include mirror clauses in this stronger sense. It includes cooperation dialogues, a TSD chapter, and the safeguard regulation discussed above (which includes a Parliament-added reciprocity provision triggering investigations on standards grounds, but only on a credible-evidence threshold whose operationalisation is unclear). The Commission's mirror-clause work and the Mercosur deal are running on parallel tracks: the deal enters provisional application on 1 May 2026; the mirror-clause assessment is still in scoping.

There is a serious counter-argument to mirror clauses, and it deserves to be heard. John Clarke, former EU head of WTO delegation, gave a 2024 interview titled "Encouraging mirror clauses is playing with fire." His argument: poorly designed mirror measures risk WTO retaliation, set precedents that other large traders could use against EU exports, and tend to be politically rushed when farmer pressure is highest, producing exactly the bad regulation that European farmers would then have to live with for decades. The Commission's caution on the topic is not pure capitulation to Mercosur producers. It reflects a real legal-trade complexity that the NGO framing sometimes elides.

That complexity does not, however, dissolve the underlying question. If the EU is committed to certain production standards — on welfare, on substances, on environmental impact — and if those standards are not applied to imports, then European producers carry costs their competitors do not carry, European consumers eat food produced under standards Europe has rejected for itself, and the standards become, slowly, less defensible at home as well. That is the structural pressure the standards gap creates, and it does not go away because mirror clauses are legally complicated.

Open questions

If the Commission's current practice on pesticide import tolerances may itself violate EU law, as the Veblen-commissioned legal opinion argues — what does that imply about how seriously the existing standards framework should be taken on its own terms?

If mirror clauses are legally complex and politically risky, but the alternative is a documented, repeated enforcement failure on substances banned forty years ago, what is the actual third option?

If the EU is willing to condition tariff access on animal welfare for shelled eggs but not for the high-volume products it actually imports — beef, poultry meat, pork, egg products — what does that pattern reveal about the political economy of European standards-setting?

And the question that runs underneath all of this: when European farmers comply with rules that imported product does not have to meet, the cost of that asymmetry is paid somewhere. Who is paying it, and who is benefiting? And is the bargain honest?

A note on process: see my AI Use Policy for how I work with AI in producing this website's content.