Investigative

The Regulator That Didn't

The Food Standards Agency was born from a scandal that killed 178 people. It was supposed to be the clean break — independent, consumer-first, answerable only to the public. Twenty-five years on, the evidence of what it became instead is sitting in plain sight. Nobody has assembled it into a single argument. Until now.

Nutriofia Investigations · August 2026

The Mandate

In March 2000, the Phillips Inquiry published its final report on BSE — the bovine spongiform encephalopathy crisis that had unfolded across the previous fifteen years. The report ran to sixteen volumes. Its essential finding was contained in a single, devastating conclusion: the government had repeatedly prioritised the economic interests of the beef industry over the health of the public. The Ministry of Agriculture, Fisheries and Food — MAFF — had been captured by the very sector it was supposed to police. It had issued reassurances it could not justify. It had managed information rather than disclosed it. By the time the link between BSE and variant Creutzfeldt-Jakob disease was confirmed in humans in 1996, 178 people in the United Kingdom were already on a path to a fatal neurological disease with no treatment and no cure.

The Food Standards Agency was created in direct response. The Food Standards Act 1999 established it as an independent body whose primary duty was, in the words of the legislation, "to protect public health from risks which may arise in connection with the consumption of food." Not to balance public health against economic growth. Not to facilitate industry competitiveness. Not to maintain the confidence of business. To protect public health. That was it. That was the whole job.

The proposal that shaped the Act came from Philip James and Geoffrey Cannon, whose 1997 independent review set out the case for an agency that was genuinely at arm's length from industry — staffed by scientists without commercial conflicts, transparent in its decision-making, and constitutionally oriented towards the consumer rather than the producer. James and Cannon were explicit about the lesson of MAFF: that a regulator embedded in the interests of its regulated industry will eventually serve those interests rather than the public's. Independence was not an optional feature. It was the entire point.

What followed was one of the most consequential examples of institutional forgetting in British public health history.

How Regulatory Capture Actually Works

Regulatory capture is rarely a scandal in the conventional sense. There are usually no brown envelopes, no secret meetings, no moment when a regulator decides to stop doing its job. The process is more gradual, more structural, and more insidious than that — which is precisely why it is so difficult to prosecute and so easy to deny.

It works through three mechanisms operating simultaneously.

The first is personnel. When the same people rotate between industry roles, advisory committee positions, and regulatory posts, the distinction between regulator and regulated slowly dissolves. The individual may be entirely sincere. But their professional networks, their intellectual frameworks, and their instinctive sense of what constitutes a reasonable position have been formed in an industry environment. They bring those assumptions into the regulatory body with them.

The second is framing. Regulatory decisions are always preceded by a question. Who defines the question determines who wins the argument. When the question is "is there sufficient evidence of harm to justify restricting this product?" the burden falls on public health. When the question is "does this product provide economic benefit?" the burden falls elsewhere. Industry lobbying — patient, well-resourced, and perpetual — works primarily at the level of framing: ensuring that the questions asked by regulators are questions that the industry's evidence is equipped to answer.

The third is inertia. A regulator that never acts on voluntary commitments that are not met, that consistently defers to "further evidence," that treats each piece of research as requiring replication before it can inform policy — such a regulator is not corrupt. It is simply slow. And in the food system, slow is a form of capture, because the industry's products continue to be sold at scale while the evidence accumulates.

The FSA has been susceptible to all three.

The Traffic Light War

In the mid-2000s, the FSA developed what it called the Multiple Traffic Light system — a front-of-pack labelling scheme that used red, amber, and green colour coding to indicate whether a product was high, medium, or low in fat, saturated fat, sugar, and salt. The logic was straightforward: a shopper picking up a ready meal should be able to see immediately and clearly whether what they were buying was nutritionally problematic. Red means stop and think. Green means proceed. The system required no numeracy, no dietary knowledge, and no time. It was designed, in other words, to work for the actual population rather than for an idealised one.

The food industry fought it comprehensively.

The British Retail Consortium and major food manufacturers rejected the traffic light system and threw their weight behind an alternative: the Guideline Daily Amount, or GDA. The GDA system displayed a percentage of the recommended daily allowance for each nutrient — expressed as a number. To know whether 22% fat was a lot or a little, you needed to understand what 100% looked like across an entire day of eating. The system placed the cognitive burden on the consumer rather than the label. It was, by any honest analysis, designed to be harder to use than traffic lights — which was the point.

The industry's campaign was sustained over nearly a decade. It funded research challenging the evidence base for traffic lights. It argued that colour-coding was stigmatising to food companies. It ran parallel lobbying operations in Brussels to prevent the European Commission from mandating the system. It succeeded. Front-of-pack nutrition labelling in the United Kingdom remains, to this day, voluntary. The current hybrid system — traffic light colours combined with GDA percentages — has been in place since 2013, but manufacturers are not required to use it. Some do. Many do not. The products most likely to benefit consumers most from clear labelling are the ones whose manufacturers have least incentive to apply it.

The FSA had a policy position. The industry had a lobbying operation. The result was a voluntary compromise that served the industry's interest — which was to prevent consumers from easily identifying unhealthy products — rather than the FSA's mandate, which was to protect public health.

The Advisory Committee Problem

Food policy in Britain is made at two levels. At the top, ministers decide. Below them, scientific advisory committees advise. The committees are the engine room of evidence-based policy — the place where the scientific literature is assessed, translated, and converted into recommendations that shape what the government does or does not do about a given food or nutrient or practice.

Who sits on those committees matters enormously. Their conflicts of interest — real, perceived, and structural — determine whose evidence gets weighted and whose does not. Their professional relationships with industry determine what questions get asked. Their institutional affiliations shape what counts as a reasonable conclusion.

Professor Erik Millstone at the University of Sussex, one of the most systematic researchers of food policy in Britain, has spent more than two decades documenting conflicts of interest across the UK's food science advisory system. The picture his research presents is consistent and troubling: the boundary between regulatory science and industry science in the British food system has been persistently, structurally blurred. Members of key advisory committees have held concurrent relationships with food industry bodies. Industry-funded research has been given equivalent weight to independently funded research in deliberations whose outcomes affect the commercial interests of the funders. The machinery by which the FSA converts science into policy has been, throughout its existence, imperfectly insulated from the commercial interests it is supposed to evaluate at arm's length.

None of this is secret. Most of it is publicly declared. The declarations are the point — they demonstrate that the problem was known and managed rather than prevented. "Managed" in practice has meant that experts with industry relationships have been permitted to participate in deliberations affecting those industries, subject to restrictions that are often procedurally complex and substantively inadequate.

Twenty Years of Voluntary Failure

In 2006, the FSA launched its salt reduction programme — a series of voluntary targets asking the food industry to reduce sodium in key product categories over a defined period. The programme was presented as a model of collaborative regulation: industry and regulator working together, setting ambitious but achievable goals, monitoring progress transparently. It was praised internationally. It influenced similar programmes in other countries.

The salt reduction programme did produce real reductions in some categories. It is the FSA's strongest argument against the charge of capture. But it is not the clean success story it is sometimes presented as. Progress has been uneven, target compliance has been inconsistent, and the programme has operated without any meaningful consequence for non-compliance, because the targets were voluntary and remain so.

The sugar story is considerably more damning. A decade after the salt programme launched, Public Health England — the successor body for nutritional policy — established a sugar reduction programme in 2016 with a target of a 20% reduction in sugar across key product categories by 2020. At the midpoint, PHE's own progress reports showed that the food industry had achieved, on average, reductions of approximately 3% in the first two years against a 5% interim target. By the time PHE was abolished in 2021 — replaced by the UK Health Security Agency and the Office for Health Improvements and Disparities — the monitoring infrastructure for these voluntary targets had been disrupted and the public accountability for the shortfall was substantially diminished.

The soft drinks industry levy — introduced in 2018 by the Treasury as a mandatory fiscal mechanism — tells a different story. Manufacturers reformulated dramatically ahead of the levy coming into force. The sugar content of soft drinks fell substantially. The contrast with voluntary reformulation is not subtle. When the consequence of failure is real — a financial penalty applied without exception — the industry moves. When the consequence is a strongly worded progress report, it does not.

The FSA and its successor bodies have known this for two decades. They have continued to recommend voluntary approaches.

The Category That Does Not Exist

In 2009, Professor Carlos Monteiro and his colleagues at the University of São Paulo published the first version of what became the NOVA food classification system — a framework that grouped foods not by their nutrient content but by the degree and nature of their industrial processing. NOVA's fourth category, ultra-processed food, described products manufactured through industrial processes using ingredients and additives not found in domestic kitchens — emulsifiers, flavour enhancers, modified starches, synthetic colours, hydrogenated fats — and typically sold in packaging designed to maximise consumption.

Over the following fifteen years, the research base around ultra-processed food became one of the fastest-growing bodies of evidence in nutritional epidemiology. Hundreds of prospective studies, systematic reviews, and meta-analyses — covering populations across dozens of countries and billions of person-years of follow-up — consistently associated higher ultra-processed food consumption with higher rates of obesity, type 2 diabetes, cardiovascular disease, colorectal cancer, depression, and all-cause mortality. A landmark umbrella review published in 2024 identified associations with thirty-two distinct health outcomes across the breadth of chronic disease.

Ultra-processed food now accounts for more than 55% of the calories consumed in the average UK adult diet. It accounts for more than 65% of the calories consumed by UK children.

There is no regulatory definition of ultra-processed food in UK law. The Food Standards Agency does not use the NOVA classification in its work. A food manufacturer in the United Kingdom may produce a product consisting almost entirely of industrial ingredients, package it with health claims, market it as a contribution to a balanced diet, and face no regulatory framework specifically designed to assess, restrict, or disclose the degree of its processing.

The industry's position is that NOVA is unscientific — a classification system based on the nature of processing rather than nutrient content, and therefore not a reliable guide to healthfulness. It is true that the mechanism by which ultra-processed food causes harm is not yet fully understood. It is also true that which specific components are harmful within UPF — the emulsifiers, the artificial sweeteners, the processing-induced structural changes, the additive combinations — remains an active area of research. The absence of a complete mechanistic account is real.

But the absence of a mechanism does not negate the associations. The evidence for UPF harm is now, in the words of the 2024 House of Lords report, "substantial and coherent." The FSA's continued failure to develop a regulatory framework for this category is not a neutral position. It is a decision — made in the context of sustained industry lobbying against the adoption of NOVA — that leaves the most damaging category of food in the British diet entirely outside the regulatory system designed to protect the public from harmful food.

The Advertising Delay

In 2021, the UK government published the Health and Care Bill, which included powers to restrict the advertising of foods high in fat, sugar, and salt before the 9pm watershed on television, and to ban paid-for online advertising of the same products at any time. The policy was presented — correctly — as long overdue. Children in the United Kingdom are exposed to more HFSS food advertising than children in most comparable nations. The evidence that advertising shapes food preferences and dietary behaviour, particularly in children, is robust. This was not a controversial scientific proposition.

The food and advertising industries lobbied against the restrictions with considerable resource and determination. The arguments rehearsed were familiar: disproportionate impact on small businesses, insufficient evidence, available alternatives, economic damage to broadcasters. The government listened. The regulations, which were due to come into force in January 2023, were delayed. Then delayed again. They eventually came into force in October 2025 — nearly three years after the original implementation date — in a form that was materially narrower than what had been consulted on.

Three years is not an administrative footnote. Three years is approximately 3,000 days during which the children's food advertising environment remained unchanged, despite legislation having been passed to change it. The companies whose products were subject to the restrictions continued to advertise them. The harm the restrictions were designed to prevent continued to accumulate. The regulatory clock ran. The advertising continued.

Parliament Noticed

In July 2024, the House of Lords Food, Diet and Obesity Committee published "Recipe for Health: A Plan to Fix Our Broken Food System." It was one of the most comprehensive parliamentary examinations of food policy in a generation. Its conclusions were unambiguous. The food environment in the United Kingdom was not the product of consumer choice. It was the product of decades of policy decisions that had consistently prioritised commercial interests over public health. Voluntary approaches had failed. The evidence for action was sufficient. What was required was mandatory intervention.

The report recommended, among other measures: extending fiscal levies beyond soft drinks to other HFSS product categories; giving the concept of ultra-processed food formal regulatory recognition; making reformulation targets mandatory and enforceable; restricting the placement of HFSS products in prominent retail positions; and — pointedly — reviewing the conflicts of interest framework governing the scientific advisory bodies that feed into food policy.

The government published its response in January 2025. It would continue to review the evidence. It was committed to a balanced approach. It recognised the importance of industry partnership. It did not commit to mandatory reformulation targets. It did not commit to a regulatory definition of ultra-processed food. It did not commit to extending fiscal levies. It did not commit to mandatory placement restrictions.

The House of Lords had assembled the evidence of a generation. The government had said it would continue to review it.

The Cost

Diet-related disease costs the National Health Service an estimated £18 billion per year. Obesity alone — one outcome of a food environment shaped by the regulatory failures described above — costs the broader UK economy an estimated £58 billion per year when NHS costs, lost productivity, and social care are included.

These are not forces of nature. They are policy outcomes. The food environment that produces them was not inevitable — it was constructed, product by product, additive by additive, advertising campaign by advertising campaign, voluntary target by missed voluntary target, over a period of decades during which a regulator was in place whose stated mandate was to prevent precisely this.

In 1996, when the link between BSE and vCJD was confirmed, the number of people who would die as a result of the regulatory failure of the preceding decade could be counted. The number was 178. It was terrible. It triggered a public inquiry. It abolished a ministry. It created an agency.

The chronic disease epidemic driven by the modern diet is not so easily counted. Its victims do not collapse from a single identifiable cause attributable to a single identifiable failure. They develop type 2 diabetes at 50. They have a heart attack at 62. They are diagnosed with colorectal cancer at 58. They spend years on medications that cost the NHS money and cost them quality of life. They die earlier than they would have if they had grown up in a different food environment. The link between their illness and the regulatory failures documented in this piece is probabilistic, not individual — which is exactly what makes it so much easier to ignore.

But ignoring it does not make it untrue. The evidence is now voluminous, consistent, and, in the assessment of the country's own upper legislative chamber, sufficient to justify mandatory action. The question of whether anything will be done is no longer a scientific question. It stopped being a scientific question some time ago.