The Labour Behind the Milk

"The Filled Dairy Loop" set out why British milk fat keeps losing ground to a legal, standing vegetable-fat alternative — falling farmgate prices, a collapsing wholesale butter market, and demand that doesn't reliably come back even when milk gets cheap. That's one side of what's squeezing British dairy. This piece is about the other side: even a farm that wants to keep producing real milk, with no interest in vegetable fat substitution, increasingly can't find people to do the work. A collapsing price and a thinning workforce are not the same problem, and it would be dishonest to collapse them into one. But they land on the same farms, in the same years, and that combination is worth setting out plainly.

What the industry's own survey says

That figure needs a plain caveat attached, not buried underneath it: it is Arla's own released survey statistic, gathered and published by a dairy co-operative with an obvious stake in the story it tells, not an independently audited academic study. It hasn't been peer-reviewed and the methodology isn't public in the way an ONS release would be. What it does have going for it is consistency — seven years running, never once dipping under 80%. A single-year figure from an interested party would be worth treating cautiously on its own. Seven years of it staying in the same narrow band, alongside independently documented farm closures, is a pattern worth taking seriously while still flagging exactly who's counting.

The closures, precisely dated

Worth being exact, because rounded phrases like "a farm every couple of days" get repeated until nobody checks them. The most recent 12 months work out closer to one every 1.9 days. The five-year figure already documented in "The Filled Dairy Loop" — 8,040 producers down to 6,850, a loss of 1,190 — works out to about one farm every 1.5 days across that longer stretch. So the medium-term rate is actually faster than "one every two days," not slower, and the most recent six months faster again. Whatever is driving this, it isn't slowing down.

What the closures are not necessarily about

It would be easy, and wrong, to read the closure numbers as proof that the labour shortage is the cause. AHDB's own explanation for the accelerated closure rate over the most recent six months is price and input costs — falling milk prices, forage shortages, rising costs of feed and energy — not a hiring problem. That's a separate, independently documented pressure, covered in full in "The Filled Dairy Loop": a 20% year-on-year fall in farmgate price and a 47% year-on-year collapse in wholesale butter price. Nothing in the AHDB producer-number data isolates how much of any single farm's exit was about a hiring gap versus the numbers simply not working. Both pressures are real and documented; treating either alone as the explanation would be worse than admitting they're tangled together.

A third, separate pressure: who inherits the farm

This isn't a labour-market fact and it shouldn't be dressed up as one. But it bears on the same question from a different angle: whether a farm has a next generation to staff it at all. Farming bodies opposed the original October 2024 version of the reform; the December 2025 revision was a significant walk-back, roughly halving the number of estates expected to be affected. Where a family farm winds down at succession rather than continuing, that's one fewer operation in the producer-number count regardless of whether anyone ever applied for a vacancy there — a different mechanism from a farm that stays open but can't fill its rota.

The migrant labour picture — and why it doesn't apply here

This is worth stating clearly because it's the easiest place for a piece like this to get sloppy. The Seasonal Worker visa route is for horticulture and poultry — picking, packing, processing on a seasonal cycle. Dairy farming runs on permanent, year-round staff, because cows need milking every day of the year, not for a six-week harvest window. If this scheme is mentioned at all in a dairy labour story, it's to show what the wider agricultural labour picture looks like — not because dairy farms draw on it. They don't, and the Arla skills-gap figure above is specifically about permanent hires.

It's also worth retiring an older claim in agricultural labour commentary: that roughly two-thirds of the farm workforce is EU-born. That traces to 2016 Labour Force Survey data, predates the end of EU free movement, and isn't current.

The genuinely open question: is this about pay, or something else?

The honest answer, on what's actually verifiable, is that both readings have real evidence behind them and neither one fully explains the figure on its own.

One secondary source, cited in industry commentary but not independently re-verified here against primary ONS or Low Pay Commission wage data, reports that dairy farmer wages rose by roughly 27% since the end of 2019 — over the same period the reported skills-shortage rate in the Arla survey rose too, from 79% to 84%. That's cited by some in the industry as evidence that pay rises alone haven't solved recruitment. Treat it as secondary-sourced, not firmly established — it hasn't been checked here against the primary wage data it would need to rest on.

Separately, the EFRA Select Committee's 2021–22 inquiry into "Labour shortages in the food and farming sector" reportedly found that employers who sharply raised wages still struggled to recruit — one example cited was a strawberry grower who raised seasonal picker pay 50% and still filled only two-thirds of roles — and recommended a long-term domestic skills and training strategy rather than treating the shortage as a pure pay problem. That finding wasn't independently re-fetched from the primary committee report here, so it's presented as reported rather than confirmed.

One Scottish dairy farmer, writing recently in a national newspaper, argued the shortage is as much about attitude as about pay — a framing some in farming hold and others reject. It's mentioned only because that debate exists and is part of the record, not because this piece takes a side in it. The claim that British workers as a group are unwilling to do this work is not one the evidence above supports, and it isn't this piece's argument.

What the evidence actually supports is narrower and less satisfying than either side's preferred story: a persistent, industry-reported skills gap that hasn't moved much in seven years, alongside wage increases that — on an unverified secondary figure — don't appear to have closed it, and a parliamentary inquiry that leaned toward training and pathways rather than pay as the more durable fix. That's not a resolved question. It shouldn't be written as one.

Two blades, one scissors

None of this says vegetable fat substitution and a hiring shortage are the same mechanism, or that either alone accounts for a farm closing every two days. They're documented separately, by different bodies, for different reasons. What can honestly be said is that they compound. A farm squeezed by a 20% farmgate price fall is less able to pay its way past a thin applicant pool. A farm that can't staff its parlour is less able to weather a bad pricing year. Add a succession system where, even after December 2025's revision, some family farms will still wind down rather than pass on, and you have three separate, independently sourced pressures landing on the same businesses in the same years — not one story, but three blades on the same scissors.

What's actually at stake

Made Properly exists to document what gets lost when a trade dies quietly — not just the businesses, but the hard-won skills that don't transfer to a spreadsheet and don't come back once the last person who held them retires. A herdsperson who can read a cow's health from how she stands in the parlour is exactly that kind of skill. So is the knowledge, built over a working life, of a particular herd and a particular land through a particular set of seasons. When a farm closes because the price collapsed, that skill goes with it. When a farm can't hire because nobody trained for the job, the same skill never gets passed on in the first place. Both routes end the same way: fewer people who know how to do this well, and a shorter list of who's left to teach the next one.

See The Filled Dairy Loop for the price side of this story, and The Adulteration Files for the full strand.