The Melamine Milk Scandal: China's Real Food Safety Conspiracy

In the spring and summer of 2008, paediatricians in scattered Chinese cities began seeing the same wrong thing: infants, some only a few months old, with kidney stones. Kidney stones are a middle-aged man’s complaint. In a baby they are almost unheard of, and here were wards filling with them — tiny patients unable to pass urine, crying inconsolably, some in acute renal failure. The doctors who first raised the alarm were told, in effect, to wait. By the time the Chinese government confirmed what was happening in September, an estimated 300,000 children had been affected, roughly 54,000 had been hospitalised, and at least six infants were dead. The cause was a white industrial chemical called melamine, and it had been added to milk on purpose, by people who understood exactly what they were defeating.

The test the fraud was built to beat

To understand the scandal you have to understand a single laboratory shortcut. The protein content of milk is not usually measured directly. It is estimated using the Kjeldahl method, developed in 1883, which measures the nitrogen in a sample and multiplies it by a fixed factor, because proteins are reliably about 16 per cent nitrogen. For a century and a half this worked because, in real food, nitrogen came from protein. The test assumed honesty in the sample.

Melamine is a nitrogen-rich compound used to make plastics, laminates and fertiliser. By weight it is about 66 per cent nitrogen — far denser in nitrogen than any protein. Stir a little melamine into watered-down milk and the Kjeldahl test reads the extra nitrogen as extra protein. The milk looks richer than it is. A dairy middleman who had diluted his milk with water to increase volume could restore its apparent protein grade for a few yuan a tonne, pass quality inspection, and pocket the difference. The fraud was not crude. It was an educated attack on the precise blind spot of the standard test — the assumption, unquestioned since the nineteenth century, that nitrogen means protein.

This is the same species of deception that runs through the horsemeat affair in Europe: find the one thing the inspection cannot see, and pour your profit into that gap. The difference is who paid. In Europe the victims lost money and trust. In China they were babies whose kidneys could not process the crystals that melamine forms in the body.

The mechanism of harm is worth understanding, because it explains why infants suffered so specifically. Melamine on its own is only mildly toxic, which is part of why the practice had persisted quietly — it did not kill adults or older children who consumed trace amounts in ordinary food. But melamine combines with a related contaminant, cyanuric acid, to form insoluble crystals, and in the concentrated, high-dose exposure of an infant fed melamine-laced formula as its sole food, day after day, those crystals build in the tiny tubules of the kidneys and block them. An adult drinking the same tainted milk occasionally would flush it; a three-month-old drinking nothing else could not. The fraud was calibrated, whether its perpetrators understood the biology or not, to fall hardest on the youngest and most dependent consumers in the entire market.

Sanlu, and how far up it went

The company at the centre was the Sanlu Group, based in Shijiazhuang in Hebei province, one of China’s largest dairy firms and a maker of cheap infant formula sold to poorer, often rural families. Sanlu was partly owned — 43 per cent — by the New Zealand dairy giant Fonterra, a detail that later mattered a great deal.

The adulteration happened upstream of Sanlu, among the network of collection stations and middlemen who bought raw milk from thousands of small farmers and sold it on to the processor. Farmers watered milk to stretch it; collection agents added melamine to hide the watering. But the crucial finding of the investigation was that Sanlu knew, and knew early. The company had received consumer complaints about sick infants as far back as December 2007. Its own testing detected melamine by around June 2008. And its response was to keep selling while it quietly tried to manage the problem — recalling some stock through back channels, negotiating with local officials, and saying nothing to the public.

The reason the delay ran for months rather than weeks brings in the second, larger institution: the state. Shijiazhuang’s municipal government later admitted it had received Sanlu’s report and sat on it. And hanging over the entire summer of 2008 was the Beijing Olympics, opening on 8 August. China had spent years and a fortune presenting a flawless face to the world, and a nationwide poisoning of infants by the dairy industry was not a story anyone in authority wanted breaking during the Games. The timeline is hard to read any other way. The public warning came in September, after the closing ceremony.

New Zealand had to make the phone call

The point where the cover-up broke is documented in unusual detail because a foreign government forced it. Fonterra, Sanlu’s part-owner, was told by its Chinese partner in early August that melamine had been found. Fonterra pressed for a full public recall and, according to its own account, was resisted by the local Chinese authorities. Frustrated, Fonterra took the matter to the New Zealand government. The New Zealand Prime Minister, Helen Clark, later said her officials, unable to move Beijing through the usual channels, went directly to the central Chinese government — bypassing the provincial administration that had been suppressing the story. Within days of that intervention, in the second week of September, the recall became public and national.

Sit with the shape of that. The mechanism that finally overrode a Chinese cover-up of a Chinese poisoning of Chinese children was a phone call from Wellington. The internal channels — company to regulator, city to province, province to centre — had all failed or stalled. It took a stakeholder outside the country, with nothing to lose politically, to make the concealment untenable. That is the recurring lesson of institutional cover-ups: they rarely break from within. They break when someone the institution cannot control gets hold of the facts.

Confession, punishment, and the fork in the story

Once it was public, the Chinese response was fast and, in places, ferocious. The Sanlu chairwoman, Tian Wenhua, was tried and sentenced to life imprisonment. Sanlu itself was declared bankrupt. Two men more directly responsible for producing and selling the melamine mixture — Zhang Yujun, who manufactured a melamine-laced “protein powder”, and Geng Jinping, who sold adulterated milk — were sentenced to death and executed in November 2009. Several local officials were sacked, and the head of China’s quality watchdog, Li Changjiang, resigned. Compared with the way corporate food scandals resolve in the West, this was punishment of a different order entirely.

The contamination also turned out to reach far beyond Sanlu and far beyond formula. When authorities finally tested across the industry in September 2008, melamine showed up in the products of nearly two dozen Chinese dairy companies, including household names such as Mengniu and Yili, which meant the adulteration had been a general practice among the collection agents feeding the whole sector, not the crime of one firm. Because Chinese milk powder is an ingredient in a vast range of exported goods — sweets, biscuits, chocolate, coffee whiteners — the recall went global: products from White Rabbit creamy candy to certain Cadbury lines made in China were pulled from shelves across Asia and beyond, and more than a dozen countries banned Chinese dairy imports outright. A fraud aimed at cheating a protein test in a provincial collection station had, through the anonymity of powdered ingredients, seeded itself into confectionery on the other side of the world.

And here the popular telling forks, in a way worth naming carefully. Outside China the scandal hardened into a simple story: Chinese food is poison, produced by a system that cares nothing for human life. That reading is understandable and partly earned, and it is also too flat to be true. The people who died for it were executed. The affected families were, eventually, offered compensation. The country overhauled its Food Safety Law in 2009, in direct response. The failure was not that no one in China cared — some of the doctors and journalists who tried to sound the alarm early cared enough to risk their careers. The failure was structural: a fragmented dairy supply of countless tiny farms and unregulated collectors, a testing regime with a century-old blind spot, and a political culture that punished bad news more reliably than it punished the thing the news was about. The instinct to bury a scandal until after the Olympics is not a uniquely Chinese vice. It is what institutions do when their reputation and the truth point in opposite directions, whether the institution is a Communist Party province or a pharmaceutical marketing department.

What it was really about

The melamine scandal is often filed as a story about Chinese governance, and it is that. But the more durable thing it exposes is what happens when a trust chain gets too long and too cheap to police, and when the incentive to look good outruns the incentive to be safe.

Consider who was poisoned. Not the middle-class families in Beijing who could afford imported formula, and not the breastfed. The victims were disproportionately the children of the rural poor, who bought the cheapest domestic formula from the least accountable end of the supply chain. Adulteration flows downhill, toward the people with the least power to demand honesty and the least capacity to detect a fraud designed by a chemist. The melamine went into the milk of the babies whose parents could complain the least loudly, and it stayed there through a summer because complaining loudly was precisely what the system was arranged to prevent.

There is a further, quieter tragedy in the aftermath. Several of the people who worked hardest to expose the scandal — parents of sickened children who organised for compensation, and campaigners who pressed for accountability beyond the executions — found themselves harassed or, in at least one prominent case, imprisoned. The activist Zhao Lianhai, whose own son was among the poisoned, was jailed in 2010 for “inciting social disorder” after organising affected families, before eventually being released on medical parole. The same instinct that had delayed the warning through the Olympic summer reasserted itself once the emergency passed: manage the story, contain the complaint, protect the image. Punishing the poisoners and silencing the parents were, in the official mind, parts of the same project of restoring order.

What lingers is not the chemistry, ingenious as it was, but the arithmetic of the delay. Somewhere in the spring of 2008, people who knew that infants were being harmed weighed that knowledge against a recall’s cost to a company, a province and a national image, and decided the harm could wait a few more months. The children in those hospital wards were the price of that arithmetic. The value of documenting a case this thoroughly — the dates, the executions, the phone call from New Zealand — is not to render a verdict on a country. It is to see, clearly and without comfort, the exact machinery by which good people and bad ones together let a poisoning run for a season rather than admit it in a week.