Snow had moved from academe to the world of government administration — the corridors of power, as he called them. He argued that recruitment into the civil service was old-fashioned because it relied too heavily on the humanities. It was partly in response to Snow that in the early 1960s Lord Fulton was commissioned to look into the state of the civil service. His report took a long time to produce but when it came out in 1968 it proposed a new way forward for civil service recruitment. The problem, according to Fulton, was that the British civil service was based on "the philosophy of the amateur, the generalist, the all-rounder". He suggested, moreover, that the service recruited in a way that went back to Thomas Macaulay's model for the Indian Civil Service in the 1830s, the essential idea being that you read classics at Oxbridge in order to prepare you to govern an empire. Fulton argued that scientists, engineers and professionals were not sufficiently represented within the civil service. He linked this to a lack of skilled management inside the corridors of power and proposed that the civil service "fast track" should have a far greater emphasis on technocrats. And especially on economists. Once recruited, high-flying recruits should then be given a crash-course in modern management practice.

It took a while for the implications of the report to sink in, and there is a long story of how the civil service only changed very gradually — the Thatcherite economies and management measures of the 1980s played a large part — but the ethos of Fulton has now triumphed. It is not politicians but civil servants who word the documents that devise the reporting requirements of the modern university. A change of government may mean a change of policy but it does not lead to a change of language. The real threat to any Newmanesque "idea of a university" comes not from the introduction of large loans or variable fees but from the pervasive influence of the language of the Treasury, of the ethos of management consultancy and of the Key Performance Indicator. What is a KIS if not a KPI?

Collini argues that this kind of language originated in free market think tanks in the 1960s and '70s. One of his recent higher education polemics is entitled From Robbins to McKinsey. In terms of the bigger picture, he is probably right: after all, the language of KPIs now pervades the NHS, the police and just about every other form of public service. But its deformation of the true function of the institution is most severe in the universities precisely because they are the places where a civilised society keeps a space apart for those things that are not immediately measurable and translatable to the language of utilitarian performance — things such as freedom of thought, "blue skies" research, critical debate, the preservation and extension of what Matthew Arnold called "the best which has been thought and said in the world" and "the turning of a stream of fresh and free thought upon our stock notions and habits".

Before the Fulton reforms, these Arnoldian values would have been in the bloodstream of the great majority of recruits to the higher echelons of the civil service. The price of Fulton's extirpation of the "all-rounder" in the name of technology, managerial efficiency and economic productivity was the erosion of those habits of thought and that care for language which sustain such values. In 1963, when Fulton was at work, the Treasury had just 19 trained economists. Something had to be done. But now, if you look at the civil service fast stream recruitment figures for the most recent year that is available, 2010, 191 general graduates were appointed out of 15,589 applicants — a 1.2 per cent success rate — and 170 economists out of 1,220 applicants — a 13.9 per cent success rate. Students graduating in economics are therefore far more likely to get into the civil service than those graduating in humanities or science disciplines. The actual recruitment numbers are even more striking than the percentages: 191 general recruits and 170 economists. This has been the pattern for a generation, so it is hardly surprising that policy is now so widely framed by the econometric model and the discourse that goes with it.