He is the Sydney Jones Professor of Education at Liverpool University, but he also happens to head up the Centre for Education and Employment (CEER). This body, for the past 20 years and from various locations, has built up an unassailable reputation for educational research.

The latest document that the Government is wrestling with is a 49-page tome entitled 'Teachers Leaving', written by the professor and colleague Dr Pamela Robinson. An even mix of quant and qual work, the study records an annual teacher resignation rate of 15.8% in 2001, 4% up on research carried out in 1999, and asks: Why is the teaching profession not renewing itself?

It was commissioned by the National Union of Teachers as part of its evidence to the Schoolteachers' Review Body, but the Union also presented it to the Government, which then became interested in various questions arising from it. It put some contracts out to tender and the CEER was successful in winning them.

So does the Government consider the qualitative elements as important as the quantitative ones? "The two are interleaved," says Professor Smithers. "Broadly speaking, there are two main ways of making sense of it all. One is through checkable evidence and the other is through narratives by which people interpret what is happening to them.

"The quantitative side is an attempt to provide a numerical map of what is going on, while the narrative fleshes it out. So you can see what percentage of primary and secondary school teachers see the workload as the primary reason for leaving the profession, while the narrative exemplifies this: 'All this is happening to me, I don't have any free time at weekends, I must leave to get a life'."

The study makes sober reading. Of every 100 final year students training to be teachers, 40 do not make it to the classroom. With the initial teacher training budget currently standing at £245m, this represents an annual loss of £100m.

Another 18% leave during their first three years of teaching, so over half the trainees ­ which should have solved the teacher shortage ­ are soon lost. The teacher cohort, aged 40 to 49, is twice the size of the present intake, with a much lower retirement rate, but retirement is already looming.

The solution is obviously to stop teachers leaving prematurely. The study interviewed 102 teachers from a cross-section of schools. They gave, typically, about three ­ overwhelmingly negative ­ reasons each for leaving, only 14.8% citing rival attractions.

As for the Centre, its composition has changed over time.These days it's a six-strong team, including an IT specialist, but at one point in its life it grew to 14. "We need people with particular skills and outlook, and at that time Pamela and I considered ourselves mainly as managers rather than doers of research," he says. "Unfortunately, the quality of the work coming in was not sufficiently high, and we found ourselves having to repeat a lot of it."

So a certain amount of downsizing went on, and a different selection process when it came to hiring researchers. "When we were relatively large we were recruiting people with M.Scs and PhDs. They knew all the techniques but, at the end of the day, didn't deliver anything which added to our understanding of the world," he says.

"We need those who have sufficient understanding and empathy with what it is they are trying to record, describe and interpret. It's a bit like being a novelist: you can explain the shape of a novel, but only a certain number of people can create something which is insightful and of interest."

Twenty years on, the CEER is beginning to sound like an institution. It is self-financing ­ even though, at times, it is difficult to line up the next commission while working on the current one. No difference here, then, from any other small business.

Its essence, however, is very simple. "The key is to retain your independence and describe it how it is," says Professor Smithers.

Let us hope that Government listens.