In the constant babble of complaints about the state of GCSEs and A-levels, the voice of employers has been the loudest and most insistent. Businesses in England and Wales have lost faith in the ability of GCSEs and A-levels, either to encourage the development of appropriate skills, or to test them adequately. That, at any rate, is what the Confederation of British Industry and the Institute of Directors have to say.

Future employment prospects are an extremely serious matter, far more serious to most pupils than the scraps of paper that certify their grades.

So employers’ views have weighed heavily in the discussion, not least in the Tomlinson committee’s attempt to re-shape the 14-19 curriculum. The existing structure of exams, concluded the former chief inspector, has failed a generation.

What then, is to be made of the following, apparently disconnected phenomena: first, that employers are increasingly resorting to psychometric and competence tests when selecting new recruits; and, second, that the high level of post-war social mobility has been declining? This surprising slow-down directly contradicts standard liberal models, which predict a straight correlation between the spread of education and the growth of a merit-based society.

Two sets of researchers have been attempting to explain these developments.

Their studies were entirely separate, but their conclusions fit together, and do so in a way which is far from encouraging, either for Mike Tomlinson, the former chief inspector, or for the students he is trying to help.

On the face of it, the growth in psychometric testing is a straightforward matter. Employers have lost trust in exams thanks to grade inflation, you might think, and must resort to their own selection mechanisms as a consequence. Certainly, this is a message broadcast by the IoD for at least the past decade and a half.

This, though, is not the picture that emerges when the actual recruitment processes are analysed. Studies at London University’s Institute of Education by Professor Alison Wolf (now at King’s London) and Andrew Jenkins have found that two main factors were responsible for the use of psychometrics. One was the gathering power of personnel departments which, perhaps, are flattered by the professional responsibility involved in administering them and are keen to see their influence grow. The second and clearest pressure for additional testing, though, came from the growth of employment legislation and regulation, the need to be seen as accountable and fair in the workplace, and the perceived difficulty of sacking anyone should the wrong decision be made.

So far, so surprising. What is of even greater significance to schools, though, is what Wolf and Jenkins did not find. There was, they say, no evidence that employers were using tests because they thought that GCSEs and A-levels were in some way a degraded currency. Although a number of them were concerned about levels of educational performance, literacy and numeracy in particular, their criticism was not directed towards the exam system, its precision or imprecision. It was more that the formal qualifications, for these employers, were beside the point. There was nothing technically “wrong” with current exams structure that the world of real businesses and employment (as opposed to the world of press statements issued by the CBI and IoD) could identify - beyond the plain fact that a sheaf of certificates appears to have given them no idea whether or not to employ a particular individual.

On this reading, conclude Wolf and Jenkins, it would be better to concentrate on the nature of the education rather than the nature of the assessment. “Employers’ declining willingness to take formal note of qualifications does call into question the idea that education and training should be as qualification-driven as they have become over the past 10 years,” conclude Wolf and Jenkins. “The unstated assumption behind much government policy seems to be that a course is worthless to the individual if no qualification is attached. This is clearly not the case if it teaches skills which can be demonstrated during recruitment testing, and it is simply untrue to imply that all qualifications have general (or even specific) labour market value.”

What, then, are employers really looking for? The second piece of research, this time from Nuffield College Oxford, brings us closer to an answer. A team led by Dr John Goldthorpe started by asking why the popularity of A-levels and the steady expansion of higher education had not brought about an increased level of social mobility. Education appears, instead, to be having a declining impact. It is a technical sounding question, but they too began with employment, the chief means by which status is gained and distributed, and at employers’ hiring practices. Again, like Wolf and Jenkins, they found there was no easy relationship between the certificates that the qualifications system was turning out, and the sort of knowledge about a person that employers seemed to want.

On the contrary, while a good professional qualification was a good predictor of success in law, engineering or some other technical occupation, other sections of the market were interested in qualities which were not even tested by A-levels or degrees. Appointments to managerial posts in the service sector depended instead on such qualities as confidence and being skilled conversationally - qualities that emerge through psychometric testing, but which, of course, also become apparent in interviews.

A large, and growing, sector of the market, then, wants the skills traditionally associated with the professional and academic middle classes.

“It’s not much use having some graceless anorak, however impressive his or her degree,” Goldthorpe commented. “The attributes that these people have from their family background have some real commercial use. It’s not nepotism. Employers know what they want.”

Suddenly the growth in personality testing takes on a new dimension. The irony is that, on this reading, what many employers are after is a simple social screening device of the sort that both A-levels and a degree provided 30 years ago. It did not matter what grade you got, what mattered was that you had been entered. Now though, what the researchers term the “noise” in the system, generated by the volume and variety of qualifications, means that such simple presumptions are impossible. The accessibility and merit-driven nature of GCSEs and A-levels is fair to students, you might say, but employers have no particular interest in being fair.

There are a number of depressing conclusions that follow. First, if the link between what employers want and what the system can turn out is broken in some way, then the Government’s ability to affect the economy through re-fashioning schools would seem to be overstated. It does not help, for example, to push people through higher education if, the more you push through, the less likely employers are to bother looking at degrees. As the authors put it: “Attempts by governments to promote educational expansion and to design educational and training policy with the aim of improving economic performance may often rest on misconceptions about the nature of modern economics.”

Nor are there grounds for expecting an “education-based meritocracy” to develop automatically, if firms remain determined to ignore evidence of technical ability - as demonstrated through, say, maths A-level - at the expense of more general personal qualities. Related to this is the implication that the continued decline in manufacturing and engineering as a mass employer - and the parallel rise of the service sector - will continue to benefit the privileged classes to the probable detriment of bright but less confident children from poorer backgrounds. Professional and technical occupations are the ones with the greatest trust in formal qualifications and the lowest use of psychometrics; this is reversed for the service sector.

It also appears that there is little Mike Tomlinson or anyone else concerned with re-configuring exam systems can do if employers are going to ignore the fruits of their labours. As Alison Wolf puts it, qualifications are “beside the point” for a range of occupations. The salient fact, she says, is the chasm between the confident, well-spoken and well-connected in British society, and the rest - a chasm that threatens to swallow up Tomlinson, just as it has the dozens of well-meaning attempts that have gone before it.

The Growth of Psychometric Testing for Selection - why has test use increased and what does this mean for education? httpcee.lse.ac.ukpublications.htmWhy Do Employers Use Selection Tests - evidence from British WorkplacesAlison Wolf and Andrew Jenkinscee.lse.ac.ukpublications.htm“Education, Employers and Class Mobility”, by Michelle Jackson, John Goldthorpe and Colin Mills, to be published in Research in SocialStratification and Mobility (Elsevier)