Government spending on translation soars by 40% in UK
Calls for tougher language tests for new immigrants have returned after government spending on translators rose by 40 per cent in two years, despite an explicit drive to cut costs.
In 2007, Labour laid out 100 million pounds on translation services across Whitehall and the justice system, then pledged to slash the bill. Last year the figure for translating information for non-English speakers reached 140 million pounds. Critics branded the spending a waste of money and demanded stricter language tests for immigrants moving to Britain.
The criminal justice system, taking in police, courts and prisons, accounted for 34 million pounds of that total. NHS Direct translates its documents into 200 languages, including Esperanto, which is spoken by roughly 2,000 people in the United Kingdom.
The Esperanto detail does a lot of work
That last fact is the one every newspaper leads with, and it is genuinely difficult to defend. Nobody in Britain speaks Esperanto and only Esperanto. It is a second language by definition, learned by enthusiasts who already have a first. Translating NHS health information into it is not language access. It is a procurement box being ticked by somebody who bought a bundle of languages without asking which ones had users.
The trouble is that the Esperanto line then gets used to discredit the other 199, and most of those are not remotely frivolous. Polish, Urdu, Bengali, Somali, Arabic, Romanian and Portuguese speakers use NHS services in large numbers, and the cost of them not understanding a discharge note or a medication label is paid in ambulances.
Where 140 million pounds actually goes
The headline figure lumps together things that behave very differently. Roughly a quarter sits inside the criminal justice system, and that money is not discretionary in any meaningful sense. A defendant who cannot understand the proceedings against him cannot receive a fair trial, and a trial that collapses because no court interpreter could be found costs far more than the interpreter would have. Adjourned hearings, re-listed cases, witnesses recalled, custody extended: the bill for absence lands in the same budget as the bill for provision, but nobody itemises it that way.
Health is the second large block. The rest is scattered across local authorities, benefits administration, immigration casework and the ordinary business of a state that publishes forms.
The language test argument
The demand for tougher testing is politically potent and analytically slippery. Britain already requires English for most settlement and citizenship routes, and a test at the border does nothing about the two groups who generate most interpreting demand: people who have a legal right to be here regardless of their English, and people whose English is fine for daily life but not for a cross-examination or a consent form.
That second category is the one policy consistently misses. Functional English collapses under stress and technicality. A man who orders his lunch, holds down a job and reads a tabloid may still be unable to follow a barrister's question about what he intended on a specific evening eight months ago. Providing an interpreter in that situation is not indulgence. It is the difference between a safe conviction and an appeal.
The efficiency question nobody asks
There is a real cost story here, and it is not about immigrants. It is about procurement. Public bodies have historically bought interpreting chaotically: hundreds of separate contracts, no shared rates, no shared quality bar, agencies charging cancellation fees for hearings that never started, and no central record of who was booked or whether they turned up. When the Ministry of Justice later tried to consolidate its interpreting into a single framework, the savings were real and the disruption was severe, with interpreters boycotting the new rates and trials collapsing for want of qualified staff.
That episode is the useful lesson. Language spending in the public sector is genuinely inefficient, but the inefficiency lives in how the state buys, not in the fact that it buys. Cut the budget without fixing the procurement and you get the same bill plus a queue of adjourned cases.
The 40 per cent needs unpacking
A rise from 100 million to 140 million pounds in two years sounds like a system out of control, and it may be. It may also be a system that finally started counting. Central government did not have a single reliable figure for interpreting spend in 2007, and much of the apparent growth in these totals reflects departments logging costs they previously buried in operational budgets. That is not an excuse. It is a warning against treating the percentage as if it measured demand.
Demand did grow, for reasons that had nothing to do with policy. Accession to the European Union brought a large Polish-speaking population into the country in a short period, and public services responded, as they were legally obliged to. Blaming translators for the consequences of free movement is a strange way to run an argument.
What good looks like
The evidence-based answer is unglamorous. Register qualified interpreters, pay them enough that the register stays full, book them centrally, measure no-shows, and stop paying for languages nobody requests. Translate the documents people actually read rather than every document in the building. And publish the numbers, because the 140 million pound figure is quoted endlessly by people who have never seen it broken down.
Until that happens, the row will repeat every couple of years, in the same words, with the same Esperanto anecdote doing all the heavy lifting. The departments that generate the bill remain the ones least willing to explain it.