We connoisseurs of administrative language like a new metaphor. The government is often willing to oblige, and in last summer's White Paper, it duly gave us "the information landscape for higher education" (paragraph 6.22).

How nice, I thought, idyllic even, to think of data as a "landscape". But there are landscapes and landscapes. In the Olympics opening ceremony, England's green and pleasant landscape morphed into Pandemonium in a matter of minutes. So if higher education information is a landscape, what kind of landscape is it?

Well, for one thing, it is vast, even panoramic. The White Paper itself notes that there are "approximately 550 lines of external reporting that institutions are asked to comply with". Fortunately, it doesn't list them all.

I too will spare you that and focus on just two of the 550: the data returns that universities have to make about their students.

³ÉÈËVRÊÓƵ

The main student return is the one made to the ³ÉÈËVRÊÓƵ Statistics Agency. Carried out by techies, its existence rarely reaches the daylight of committees or external media, save when things go badly wrong. But it is a major undertaking, requiring a significant investment of a university's resources.

This is accompanied by the ³ÉÈËVRÊÓƵ Students Early Statistics return, which is sent directly to the funding council. This is a snapshot of student numbers in December of each academic year that exists primarily to drive funding and student number control.

³ÉÈËVRÊÓƵ

The Hesa return is produced the following October. One of its functions is to verify the HESES return: for example, were the predictions in the HESES return about the number of students who would not complete their courses borne out by later events? Thus the existence of two surveys creates a third exercise, namely the need to reconcile them.

I do not wish to suggest that this is a complete waste of time and effort. Both exercises produce a pool of data from which politicians, the media, researchers and universities themselves can fish. They can show how the student body changes over time, which students drop out, and so on. In time, they will provide the main evidence for whether the new tuition fees regime has or has not increased recruitment from disadvantaged socio-economic groups. But their cost and cost-effectiveness are rarely subject to scrutiny.

The London School of Economics' Hesa return is, I imagine, more straightforward than most. We have a limited range of subjects, we work to a single academic year start date, we don't offer sub-degree provision, we don't do teacher training, we aren't Welsh (who are subject to special requirements). Even so, when we last counted, our return contained well over a million different bits of data. As landscapes go, the Hesa and HESES student returns are densely populated.

To identify conflicts in data, or even just errors, the return is governed by a set of so-called "business rules". There are 654 of them at present. To make sure that producing the return never becomes dull, Hesa changes the compilation processes or the codes from year to year. So last year, universities had to address new coding for "highest qualification on entry", new coding for disability, changes with respect to the 13-month rule for non-completions (I suggest you don't ask) and new codes for collaborative programmes. It took us an email exchange of 10 messages to be told that we didn't have to apply these new codes to our own collaborative provision. "There is an inconsistency in the guidance and this is something we are looking to align with the ³ÉÈËVRÊÓƵ Funding Council for England's assistance in the coming years," we were told. That's good to know.

More recently, Hefce itself has decided to get involved in the Hesa return, in the name of judging whether the data are fit for purpose. In principle, this complements rather than duplicates Hesa's checking. But this new arrangement started last year with Hefce asking us questions that Hesa had already asked.

The landscape of Hesa returns is full of odd, distorted perspectives. Language fragments. It is a world in which a year can be long or short. A student can be an "instance", or even two "instances" within the same year. Following recent correspondence with Hesa, we now include in our return "institution-owned halls" that we don't actually own.

I once chaired a meeting where a senior officer from the funding council spent 45 minutes explaining to about 40 academic registrars what was meant by the word "non-completion" for the purposes of the Hesa return. Those who at the start of the meeting thought they knew the answer certainly didn't by the end.

Here is a message that Hesa sent round on an administrators' mailbase earlier this year:

³ÉÈËVRÊÓƵ

"Broadly, you are right the change to the counting method in HESES11 has led to a consequential change in the year to which FUNDCOMP will relate for students on non-standard academic years, as the guidance states it relates to the year of instance being funded. Thus, for a student on a non-standard year of instance, the FUNDCOMP will normally be 3 in every year, except their last year. Where this will not be the case is where it is already known that the student will not complete the non-standard year that they have started, even though they may not have withdrawn. For example, if a student started their year of instance in January 2012 and failed to complete a module in May 2012 and there is no option for them to recover that completion before the Feb 2013 then code 2 would be returned, even if they had not withdrawn by 31 July. I hope this is clear."

³ÉÈËVRÊÓƵ

Perfectly, thanks.

But, like many landscapes, this may be a landscape under threat. Paragraph 6.21 of the White Paper was categorical. "In the short term, we will work with the ³ÉÈËVRÊÓƵ Statistics Agency to reduce the size of data collections through the periodical review process." This is stirring stuff.

What the government means by phrases such as "short term", however, is never entirely clear. In the 14 months since the White Paper was published, Hesa has added new fields and data items - in part to meet government requirements - and has taken nothing away.

Instead, the sector's attention has focused on a more modest recommendation in the next paragraph, that various bodies work together "to redesign the information landscape for higher education in order to arrive at a new system that meets the needs of a wider group of users, reduces the duplication that currently exists, and results in timelier and more relevant data".

This challenge has been taken up enthusiastically by the Interim Regulatory Partnership Group: Project B, which produced a report earlier this year. It runs to 43 pages, with some splendid graphics.

The main recommendation is for the key stakeholders "to establish a collective oversight of the information landscape, to achieve a more efficient and effective system of governance. This would enable a programme of work, using shared expertise and building on the key strengths identified in the sector, to create a more coherent set of arrangements for the collection, sharing and dissemination of data and information." That, too, is perfectly clear.

There is a feeling of d¨¦j¨¤ entendu about language of this kind. We last used it in 2008 or 2009, when we talked at various techie gatherings about engaging with the MIAP initiative - Managing Information Across Partners. Just to reinforce this sense of wheels turning, that earlier initiative was itself set up by the government in 2002 "to reduce bureaucracy across the sector by improving the management of information".

Project B proposes to develop "a calendar and inventory of data collections across the year" and "a data model, lexicon and thesaurus for the sector". The calendar is a "first step" towards "streamlining collections". But the ways and means to reducing duplication remain vague. The recommendations could even increase the burden on institutions, at least in the short term, before any of the supposed benefits kick in.

There are other, more immediate ways of tackling this. For example, Hesa and Hefce could look at stripping back the need to return details of a student's activity, especially at module (course) level. At the macro level, the government could instruct its agencies to rethink the HESES return, perhaps by limiting it to institutions that, under the new financial regime, will still receive significant amounts of funding for teaching or are still subject to serious student number controls.

As a final thought, I want to suggest that discussion of these student returns is part of something wider. Allow me to return to the landscape metaphor for a moment. Sidney Keyes, who died in 1943 at the age of 20, wrote a poem titled Sour Land about the part of Oxfordshire where Alexander Pope was living when he wrote part of his translation of the Iliad. I have been there. There are nicer parts of Oxfordshire. Joining the geography and Pope's poetic activity together, Keyes wrote: "This landscape of bulbous elm and stubble/Sharpens the mind into revolt at last."

Our minds in the higher education sector need to be sharpened into revolt as well. Every day we see signs that the free market is getting closer. But the associated freedoms for universities, especially freedoms from regulation and control, stay stubbornly at a distance.

The government has made some breezy promises, especially in the White Paper. But this is the same government that, in the name of improving information, has introduced the Key Information Set - a somewhat incoherent data set, though that may be the subject for another article. It has introduced the related Wider Information Set, which specifies and categorises the information universities should publish to students and "the wider public". It is also thinking about introducing national surveys and key information sets at taught postgraduate level. Having made these promises, the government does not seem to be pushing them on its agencies especially forcefully.

³ÉÈËVRÊÓƵ

The government also talks from time to time about adjusting the ratio between academic costs and administrative costs in favour of the former. If it is serious about this, deregulation is a good place to start. New metaphors are nice; kept promises are even nicer.