The Adulteration Files: What British Food Law Actually Permits

A bar of milk chocolate can legally be 5% vegetable fat instead of cocoa butter. A tub can be sold as ice cream while containing no cream, no fresh milk, and no dairy fat at all. Neither of those facts describes a scandal. Both describe the law working exactly as written.

That's the premise of this strand. Made Properly has spent nine sectors documenting the British manufacturers still doing things the traditional way — the cheesemakers, the smokehouses, the family firms that never took the cheaper route. The Adulteration Files asks the question that follows naturally from that work: what does the cheaper route actually look like, and who wrote the rules that allow it?

What this strand is, and what it isn't

This is not an exposé of criminal behaviour. Nothing in these six articles alleges that a named company breaks the law. The opposite is closer to the truth: the entire strand exists because the substitutions it describes are legal, disclosed in the small print, and have been for decades. The story isn't that anyone got caught. It's that the rules were written to allow it, and almost nobody reads far enough down the ingredients list to notice.

That distinction matters because the temptation, writing about food, is to reach for alarming language and let the reader do the rest of the work. This strand doesn't do that. Every substantive claim below carries a visible grade — Documented Law, Hard Data, Peer-Reviewed Evidence, or Contested Hypothesis — so a reader, or a hostile expert, can check exactly how much weight each sentence is meant to carry. If a claim doesn't fit one of those four tiers, it doesn't get published. There is no fifth badge for "we think."

How to read the badges

A Documented Law badge means we have gone to the statute itself, on legislation.gov.uk, and quoted the operative regulation or schedule. If we say the law permits something, you can click through and read the same clause we read.

A Hard Data badge means an official or industry statistics body — AHDB, DEFRA, ONS, EFSA, the FSA — publishing a dated figure. Undated numbers don't run. A price or a count without a date attached is not evidence; it's a rumour with decimal places.

A Peer-Reviewed Evidence badge names the journal, the year, the cohort size, and the effect size with its confidence interval. Where the study is observational — which is most nutrition research — we say so explicitly, every time, because association is not causation and pretending otherwise is how good research gets misused.

A Contested Hypothesis badge names the paper that proposed the idea and states, plainly, where the idea currently stands with the regulators and professional bodies who would recognise it if the evidence justified that. This tier exists because live science is genuinely interesting and readers deserve to see it — but only labelled as exactly what it is: a hypothesis, not a finding.

Why the gap between label and contents is legal by design

Food law in Britain does two things at once, and it's worth being honest about both. It sets floors — minimum standards a product must meet to use a given name — and it sets exemptions, the specific, bounded circumstances in which a manufacturer can substitute a cheaper ingredient and still use that name. The floors are genuinely protective. The exemptions are where this strand spends most of its time, because the exemptions are where the gap between what a label implies and what a product contains gets legally opened.

Two patterns recur across the six investigations that follow.

The first is substitution within a tolerance. Chocolate is the clean example: up to 5% of the finished product can be vegetable fat other than cocoa butter, and the product still gets to call itself milk chocolate. Nobody is hiding this — it's in the regulation, in a schedule most people will never open. But the gap between "chocolate" as a word and "chocolate" as a regulated category with a 5% tolerance built in is exactly the gap this strand exists to describe.

The second, more striking pattern is standards that used to exist and don't anymore. Ice cream is the case study. Britain once had a legal compositional definition of "Dairy Ice Cream" — a minimum fat percentage, and a requirement that the fat be exclusively milk fat. That definition was revoked between 2014 and 2018. Nobody replaced it. The result isn't that ice cream got worse overnight; it's that the legal floor beneath a phrase millions of people still trust simply isn't there anymore, and almost nobody noticed it go.

Both patterns share a feature worth naming directly: they are legal, and describing them accurately does not require alleging that anyone acted in bad faith. A regulation permitting 5% substitution was written by a regulator, for reasons a regulator can explain. A standard getting revoked in a deregulatory sweep is a policy choice, made in public, that simply went unremarked. This strand states the rule and the documented economic incentive behind it, and leaves the reader to draw their own conclusion about motive. That's a deliberate choice, not a hedge — allegations of intent are the fastest way to turn a well-sourced piece into an easy target, and they add nothing a reader actually needs.

How this strand names companies

Made Properly has always named companies — the directory exists to name the firms doing it properly, and naming firms doing it differently is the other half of the same project. But naming a company carries risk, and this strand only takes that risk on ground that cannot be shaken from underneath it.

We will name a company when we can point to a documented compositional fact — the ingredients list as published by the manufacturer or shown on a retailer's product page, with the date it was checked — or a documented legal fact, meaning which reserved description a product does or doesn't use, and what the regulation actually requires of that description. We will name ownership and acquisition dates where they're published in company accounts, and we will quote both sides of a formulation where the same brand is made differently in two countries.

We will not do the opposite. We will not assert or imply that a named product harms anyone's health — that claim doesn't belong in a piece about labelling law, and dressing it up as one would be dishonest about what the evidence actually shows. We will not run a brand claim whose only source is a forum thread, a social media post, or an uncited blog — if the only source is a Reddit thread, for the purposes of this strand, the claim does not exist. We will not allege that a company knows a substitution harms someone and does it anyway; we'll state the rule and the documented commercial incentive, and leave the inference to you. And we will not publish an undated ingredients claim, because formulations change and a stale claim quietly becomes a false one.

One more boundary, inherited from the site's existing rules for the heritage directory: a brand owned by a conglomerate or private equity firm can never appear as a directory entry — it can only ever appear here, as evidence. And no independent firm already in the directory gets implicated by proximity to a story like this one. We check that adjacency before anything runs.

The five investigations ahead

The Five Percent Rule — the chocolate regulation in full: what Schedule 2 permits, which fats qualify, and why the same bar is made differently in the United States. This piece stays on quality and transparency ground only, because it feeds directly into an active campaign — see below.

The Ice Cream You Think You're Buying — the strand's centrepiece. Britain wrote down, in law, what "Dairy Ice Cream" had to mean. Then, between 2014 and 2018, it quietly took the definition away. What's sold under that name now rests on habit and reputation, not statute.

The Filled Dairy Loop — the economics behind the substitution. British dairy farms are closing at a rate of roughly one in seven over five years, farmgate and butter prices have collapsed, and manufactured food increasingly reaches for a cheaper fat instead. The connection between those facts is the story.

Food and the Brain: What the Evidence Says — the health chapter, handled the way the rest of this strand insists on: three real cohort studies with their real numbers, a named contested hypothesis correctly labelled as contested, and a section on what the evidence does not show, including two popular claims this strand actively corrects.

How to Read a Food Label — the practical companion to everything above. Fewer words, more decoders: what specific phrases on a UK label are legally required to mean, and which ones mean almost nothing at all.

The steel-man

None of this works as an honest strand if it only ever makes the prosecution's case, so here is the defence, stated as strongly as we can make it.

Supermarkets and large manufacturers make food affordable and safe at a scale no farmers' market or artisan producer could match, and the overwhelming majority of British households depend on that scale every week. Vegetable fats are not inherently dangerous — the argument here is about substitution, disclosure and price, never about safety, and conflating the two would be dishonest. Palm oil in particular has a real environmental case in its favour: it produces more oil per hectare than any competing crop, which means the land-use footprint of avoiding it entirely is not automatically smaller. And observational nutrition research — which underpins most of what this strand can say about diet and long-term health — is heavily confounded by income, education and lifestyle. Every study cited in this strand says so, because it's true and because pretending otherwise would make the whole exercise less trustworthy, not more.

Why this, and why now

This site started by cataloguing objects — pottery, shoes, steel, leather — because objects make the argument for themselves. Pick up a hand-forged pair of scissors next to a stamped one and you can feel the difference in your hand. Food doesn't offer that shortcut. Nobody can feel the difference between a chocolate bar with cocoa butter and one with 5% palm oil substituted in; the sensory gap is small enough that most people will never notice it, which is precisely why the legal gap matters more, not less. The ninth sector, Real Food, established that honest food production is getting scarcer in Britain for the same structural reasons as honest manufacturing. This strand is the natural next question: scarcer compared to what, exactly, and who wrote the rules that let the difference go unlabelled? The health chapter later in this strand takes that question one step further, into what the diet those substitutions produce is associated with over a lifetime — carefully, with every caveat the evidence demands, because that's the only way the question is worth asking in public.

The reader's takeaway

You don't need to memorise five pieces of legislation to use this strand. You need one habit: read past the front of the pack. The name on the front of a product is doing marketing work. The ingredients list on the back, and occasionally a schedule of a statutory instrument neither of us will ever be asked to recite, is doing the legal work. Where those two things diverge, the ingredients list is telling the truth and the front of the pack is telling you what the law allows it to imply.

That's not a reason to distrust every product in your kitchen. It's a reason to know which five or six words on a label are actually protected, and to spend the extra thirty seconds checking for them. The five investigations that follow tell you exactly which words those are — and, in the case of ice cream, which words used to be protected and no longer are, which is a different and more useful thing to know.