A couple of years ago, the University of Nottingham introduced an MSc in chemistry and entrepreneurship. Introduction of the course, run jointly with Nottingham University Business School, followed a dramatic slump in the numbers studying chemistry at universities across the UK, with the number accepted on to degree courses falling to just over 3,000 in 2003 from a peak of more than 4,300 six years earlier. The slump had already caused departments in some institutions - Queen Mary, University of London, and the universities of Kent and Exeter among others - to close.

Linking the subject with entrepreneurship was a clever idea. It tapped into the growing popularity of business studies, into government efforts to encourage a more entrepreneurial workforce, into trends towards interdisciplinarity and into increasing interest among students - particularly since the introduction of top-up fees - in studying something directly related to a future career.

The blurb for the degree makes no bones about it and promises to give students the technological and business background that will allow them "to make a significant contribution to today's chemistry-based, technology-driven economy".

Since then, Nottingham has added a number of other courses along the same lines: crop biotechnology and entrepreneurship; computer science and entrepreneurship; sustainable energy and entrepreneurship; information technology and entrepreneurship; and low-carbon technologies and entrepreneurship.

ÖйúAƬ

What subjects people choose to study - and institutions choose to offer - has always been dependent on the changing priorities of the world beyond the campus. If they had not, students would still be studying grammar, rhetoric and logic rather than media studies. But this process has accelerated in recent years as the pace of technological change has picked up, and as ÖйúAƬ has become increasingly responsive to student demands.

This has raised questions about the sort of courses that universities offer and whether they are geared to producing graduates with the skills needed by the economy, or whether the courses are being set up in response to student demand that can be driven by external factors such as popular television programmes. But just as an increase in the popularity of a subject can lead to the setting up of new courses, a decline in demand can lead to courses being closed down.

ÖйúAƬ

The ÖйúAƬ Funding Council for England responded to this challenge in 2005 by creating a programme of support for subjects deemed to be strategically important to the economy and to the ÖйúAƬ knowledge base. Hefce continues to monitor subject provision.

More recently, the Scottish Government established a task force to look at the future of ÖйúAƬ funding in Scotland. Among the proposals being considered is greater central planning of student numbers on courses.

"The simple fact is that it's a buyers' market," says Ian Jones, market research manager at Staffordshire University, which in recent years has introduced a number of innovative degrees, including forensic science and music technology, in response to demand.

"Technology is moving on, customers' requirements are becoming more sophisticated and now they are in a position to choose. Customers are in a much stronger position to find a product or course that meets their needs," he says.

In the most recent Patterns of ÖйúAƬ Institutions in the UK report, produced for Universities UK last year, Geoffrey Crossick, chairman of UUK's long-term strategy group, referred to "significant changes (over the decade from 1996-97 to 2005-06) in the subjects that students are studying".

The report highlights a rise in enrolments in subjects allied to medicine (although much of this was accounted for by the shift of the funding of nursing courses into ÖйúAƬ); biological sciences (mainly microbiology and psychology); mathematical sciences; computer science; law; mass communication and documentation; historical and philosophical studies; and creative arts and design.

By contrast, enrolments in architecture, building and planning, agriculture, the physical sciences, and engineering and technology failed to rise at the same rate as overall enrolments.

Latest figures from the Universities and Colleges Admissions Service reinforce these trends, with law taking the top position, pre-clinical medicine the second, followed by psychology, English and management studies.

ÖйúAƬ

Brian Ramsden, who produced the Patterns report, notes a general shift away from conventional subject choices - that is, subjects studied in school and taken at A level. Although English remains popular, sports science, cinematics and drama, for example, have been booming.

David Roberts, managing partner of the ÖйúAƬ consultancy the Knowledge Partnership, suggests that one reason for this is likely to be the shift away from conventional students - that is, those who come straight from an A-level course, whose parents have attended university before them, and who pursue full-time academic study for its own sake. As ÖйúAƬ has expanded from an elite to a mass system, subject choices have become driven less by academic considerations and more by what they might lead to beyond academia.

"We have gone from a situation where people did English literature because they were interested in it and good at it to one where people are increasingly seeing ÖйúAƬ as a means to an end," he says.

This means that what happens at university has to be more sensitive than in the past not only to real employment and lifestyle patterns, but also to how they are perceived, and to how people think they will be perceived in the future.

What drives these perceptions is complex. It starts in school. Although, as Ramsden points out, the link between A-level topics and choice of degree subject is looser than it was, school is still the place where students develop learning habits and enthusiasms.

If a subject is not offered at school, or is badly taught, pupils won't find out whether or not they like it enough to continue studying it for another three or four years. Changes to the maths curriculum in 2000 were blamed for a slump in the number of students applying for mathematics degrees, although the figures are now rising again.

Meanwhile, concerns have been raised about the Government's recent policy to allow pupils to drop a modern language after the age of 14. Many university teachers of modern languages, a subject that has already seen a significant decline in popularity in recent years, fear that this will reduce further their pool of potential applicants.

Such a decline can become self-perpetuating. A dearth of graduates in a particular subject means that those with the right qualifications are snapped up into lucrative jobs, leaving fewer to go into lower-paid teaching jobs. This in turn means fewer well-qualified and inspiring teachers and consequently fewer people wanting to study the subject at university.

In addition, for certain degree courses, such as medicine, pupils can apply only if they have studied the requisite A levels, so the kind of advice offered at school is crucial. But, in some cases, students who have studied the right subjects may find that the A level and university curricula bear little relationship to each other. While they may have loved the subject at school, students may see little to engage them in the same subject offered at a higher level.

Again, in modern languages, some have argued that one reason for the decline in university applications was the fact that the school curriculum had become more skills based while university degree courses were more about studying literature.

Roberts says that the alignment between how subjects are taught at school and university has become more important because students are looking more closely at exactly what is involved in a university course - a development boosted by the ease of web searching.

"Increasingly, we have found that students are interested in the detail, especially where there are a lot of programmes to choose from," Roberts says. This means that if they are interested in a subject such as law or business studies they will look at the individual modules offered as well as modes of assessment and work placements before making their choice. And it isn't just students who are looking, he says. Since the advent of tuition fees, parents have become more involved than they used to be and will nudge their offspring into subject choices they feel will serve them well in the job market.

David Watson, professor of ÖйúAƬ management at the Institute of Education, says that professional requirements and opportunities are increasingly driving changes in subject choices. This explains the growing popularity of social work, nursing and media studies. The last, in spite of reservations expressed in the media, has a high employment rate. As students face leaving university with a substantial debt, they are becoming more focused on what sort of degree courses will lead to jobs that will enable them to pay their loans back.

But the link between career opportunities and student subject choice is not always straightforward.

Some of the sciences, which have excellent employment records, still struggle to recruit. One reason is that students seem to be more influenced than ever before by how difficult they perceive a subject to be. Those choosing a course with one eye on their CVs are unwilling to risk studying for a degree in which they may struggle to secure a first or an upper second. This is one explanation given for recent declines in the popularity of chemistry and physics.

Then there's television. Veterinary science shot up in popularity after the docu-soaps Vets' School and Vets in Practice. More recently, forensic science has been booming following programmes such as Silent Witness and CSI. Ucas now lists 264 forensic science courses, including "forensic science and film studies" and "forensic science and tourism". The latest Ucas figures show that 1,738 students were accepted on forensic and archaeological science courses last year, a slight drop on the previous year when the figure was 1,878, but a significant rise from the 1,488 students accepted in 2003.

What seems to inspire students is not just the fact that a subject has featured on television but that it has featured in an appealing employment context.

While physicists and engineers rarely reach the small screen in any guise beyond that of a white-coated boffin, students can picture themselves putting to practical use all that forensic science theory they will be picking up in college.

However fanciful this may appear, Watson argues that student interests are often more rational than people would like to credit. They proved to be ahead of universities in predicting the rise of both the service and the creative economies, for example.

But they are not the only ones influencing what's in and what's out in terms of subjects studied at university. Employers also pay an increasingly important role, particularly since the publication of Lord Leitch's report in 2006, which urged universities and employers to work more closely together to improve the UK's skills base.

Gill Howland, executive pro vice-chancellor at Staffordshire University, says the institution's decision to offer business psychology has been driven by employer interest, as have moves to offer courses combining leadership and management experience with knowledge of particular areas, such as health or engineering. Her university is working hard to develop methods of responding as quickly as possible to employers' changing demands without compromising quality.

An added complication for institutions is that they need to think about not only what UK employers want but what is going on overseas as well. This is not only because many home students will go on to work abroad, but also because institutions want to attract international students.

Students from different countries make different subject choices, often related to employment opportunities in their home countries. Over the past few years, the number of home students studying electronic engineering, for example, has been dwarfed by the number from overseas. Similarly, many business and computing courses are dominated by students from China and South-East Asia.

The difficulty for institutions is that they need to be sensitive both to changing fashions in foreign employment markets - tricky when the employers are on the other side of the world - and to changes in the number of students from certain countries choosing to study in the UK.

ÖйúAƬ

This is becoming ever more difficult as the global ÖйúAƬ market becomes more competitive. While some institutions have responded by providing courses specifically aimed at the international market - for example, London Metropolitan University's masters in international human resource management - the danger is that this leaves them vulnerable should a more enticing course be set up anywhere else in the world, particularly if the course is dominated by a single nationality.

What adds to the risk is the possibility of a single unforeseen event - disease, war, a change of government policy in a particular country, new visa restrictions - causing student numbers to dry up. In the late 1990s, a financial crisis in South-East Asia caused a dramatic decline in the number of students coming to Britain from Malaysia, with knock-on effects for some of the science-based courses favoured by those students.

The effects of such events apply to courses popular with home students too. Civil engineering benefited from the construction of the Channel Tunnel in the 1980s. The Olympic Games in London are expected to boost both engineering and tourism courses.

Events and longer-term developments at home and abroad influence not only the agendas of students and employers but also that of the Government. The recent rise in the Chinese and Indian economies, for example, has made politicians anxious to improve the scientific and language skills of UK workers to help the UK compete in terms of trade and productivity.

In 2005, the Government launched a programme to help "strategic and vulnerable subjects" of national importance and last February announced that it was putting an extra ?160 million into promoting and supporting specific subjects such as science, technology, engineering and mathematics, area studies and languages.

This followed the closure and threatened closure of subjects such as physics, chemistry and engineering due to the cost to universities of providing them in light of declining applications. It also reflected concerns that some subjects, such as languages, were becoming less popular with students, not least because of a decision to downgrade languages education in schools.

The funding council's support is designed to increase demand for sciences and languages among schoolchildren and also to prevent the closure of departments until such time as demand picks up. Hefce's official line is that there is no cause for "substantially greater intervention" in course provision than there is at present. However, the council is acutely aware of the potential for a demand-driven ÖйúAƬ market to unbalance and distort course provision. To this end it is now monitoring the supply and demand of ÖйúAƬ courses on an annual basis.

The Scottish task force, made up of universities and the Scottish Government, is looking at how ÖйúAƬ can make a greater contribution to the Scottish economy. One proposal under consideration is to tightly control student numbers in particular subject areas, based on labour market projections.

A Scottish Government spokesman says: "At present, some courses such as medicine already have their numbers controlled, based on projections of future workforce demand. While we have no plans to extend this approach to other courses ... we need to consider the merits and disadvantages of any future change in this area."

Many recognise the need for a degree of government involvement in deciding what subjects universities offer.

Richard Pike, chief executive of the Royal Society of Chemistry, says that one of the concerns when chemistry departments began closing was that the subject was no longer available in large regions of the country - particularly worrying since more students were choosing to study at university while living at home. He urges a longer-term approach to subject planning and wants more co-ordination between science and education, and between science at school and university levels.

UUK has welcomed the additional funding for high-cost and vulnerable subjects, as well as efforts by Hefce to stimulate demand in physics, chemistry, mathematics and computer science.

But a UUK spokesman says: "Universities are autonomous institutions and we strongly believe that they need the freedom to make important decisions about course provision based on their own circumstances." He adds that it is recognised that a key part of the problem with science, technology, engineering and mathematics subject provision is lack of student demand.

"Universities have been very effective at responding to change, despite being under continuous financial pressures," the UUK spokesman says. "This has included collaboration, restructuring and pursuing innovation in course provision. Institutional autonomy and freedom has been integral to this success."

Even Lord Sainsbury, who was Science Minister when the subjects of strategic national importance programme was announced, said: "I hope we never tell universities what departments they should run."

Fiona Martland, executive secretary of the Engineering Professors' Council, says that while the Government needs to protect core science subjects, it is also essential for a discipline such as engineering to develop new areas, such as nanotechnology.

Pike, too, recognises that popular new courses such as forensic science have their place. While they may not deliver the depth that some employers want, the fact that they are attracting people to science is to be welcomed, he believes.

Certainly, the idea of extending central government control of universities' curricula runs counter to the more consumer-led model that most universities now follow, although being responsive to the vagaries of subject fashion does have its problems.

John Callen, director of Callen Associates, a consultancy that has recently been looking at the role of marketing in university corporate planning, warns of the dangers of "portfolio creep" - or endlessly extending the number of different courses on offer in a university portfolio but diluting the student experience (and depleting university coffers) in the process.

Then there is the temptation simply to repackage what is already on offer - calling chemistry forensic science, for example. "It is very tempting for someone who has an unpopular subject area to dress it up, give it a sexy new title and hope the punters will buy it," says Jones. "It's not the way to create a sustainable business."

He says that institutions have to be able to respond quickly and flexibly to changing customer needs but then consolidate their position. While a course will attract students through novelty value in the early years, its ability to continue attracting them will depend on reputation.

And this is something more vulnerable than ever before, thanks to chat rooms, social networking sites and the National Student Survey, all of which allow students to see whether a course that sounds interesting lives up to its billing.

Roberts warns that institutions have to think of building reputation in a number of markets. "Something that looks sexy to a 17-year-old may look a bit superficial to an employer," he says. There is also the time lag to consider. Given that it takes a year or two to set up a course, and three or four years before the first cohort graduates, what looked trendy and contemporary when a student signed up can be dated by the time he or she is looking for a job.

This time lag can cause practical problems too. If institutions really want to offer something new and substantial, they will need to employ new staff and decide what to do with existing faculty, but there is always the possibility that fashions may revert in a few years. The University of Central Lancashire decided to reopen its chemistry department five years after it closed.

Cullen says that one of the things quietly being talked about by university managers is the fact that academics will need to be more flexible, and will need to be supported in becoming more flexible, so that they can be deployed wherever demand arises. But, he says, this will still work only within broad areas of academic expertise.

Another way of achieving flexibility, suggests Roberts, is to share course architecture across a number of disciplines, perhaps developing a new course around a core subject made more contemporary and exciting with supplementary cross-disciplinary modules. But a danger here is that a new course that shares much of its architecture with another programme risks competing with that programme. Rather than increasing the total number of applicants, this may lead students to simply apply to both courses.

What everyone is in agreement over is that quality is key. However fickle students, employers and even governments are in their enthusiasm for different subjects at different times, and regardless of whether institutions are designing courses for students or "customers", the need for a good product at the end of it remains constant.

CHANGING TASTES: THE RISE AND FALL OF COMPUTING

Courses in computing started in the 1970s, reached their peak in the 1990s and declined in the early part of this century. The number of students accepted on to computer science courses has been going down steadily over the past five years.

While computing remains in the top ten most popular subjects for undergraduate study, the number of applicants accepted last year was 11,682 compared with 15,491 in 2003, according to figures from the Universities and Colleges Admissions Service.

Stephen Hagan, a lecturer in the School of Computing and Mathematics at the University of Ulster and manager of the ÖйúAƬ Subject Centre in information and computer sciences, says that the popularity of the subject depends on how students view the jobs market.

This means that universities with clear local employment routes in computing are recruiting well. Others may have to make do with smaller cohorts of students or branch out more into creative technology, music, games and other forms of multimedia, he suggests.

Confidence in gaining employment in the sector was shaken a few years ago when the dot-com bubble burst, although Hagan says there will always be demand for good computing science students and increasingly for students who are good at integrating systems.

"We try to make students understand that it is a profession, that there are jobs available," he says.

Hagan wants to see computing curricula kept more up to date, with emphasis on the applied side of the subject and how it impacts on students' lives, such as through mobile phones and social networking sites.

ÖйúAƬ

He also argues for the need to encourage more women "by showing them that it isn't a nerdy science".