In the average secondary school, the equivalent of one teacher in four is paid not to teach at all. So class sizes could be brought down significantly, and without any additional funding, simply by putting those teachers back to work in the classroom.

Around 80 per cent of schools’ budgets go on staffing costs, with the salaries of the teachers taking at least 70 per cent. Yet similar schools can vary by at least 5 per cent in total spending on teachers.

One school spends only 70p in every budget Pounds 1 on salaries and its neighbour down the road may spend 75p for an identical numbers of pupils. But such comparisons make an assumption worth questioning: that the money for teachers is actually being spent in the classroom.

Non-teaching periods are regarded within secondary schools as essential for a variety of reasons. Expecting every teacher to teach a full timetable of classes would burn out the staff very quickly, and leave none to cover absent colleagues.

For the head grappling with a rising pupil roll and a falling level of annual income, one management alternative to larger classes might appear to be a reduction in that level of non-contact time. After all, in a Group 6 comprehensive, every additional period taught by the existing staff would be the equivalent of 1.5 more teachers at no extra salary cost.

But how can the new head make sense of the pattern and any flexibility in the school they inherit? What level of classroom contact is reasonable?

One answer is to network with similar schools to share management information. For the last two years, the heads of six large 11-18 comprehensives have been doing just that. Beginning with budget figures, the exchange of information and ideas has increasingly focused on managing resources.

Schools vary enormously in how they organise the teaching day and the length of the lesson used. The timetable models in the six comprehensives range from various weekly patterns of 35 to 80 minute lessons, to a l0-day pattern of seventy 40-minute periods. It is impossible to find the comparative baseline for teaching load simply from how many teaching periods are allocated to classroom teachers.

However, the complications of each school’s timetable can be avoided if teaching time is expressed as a percentage of the total contact possible for each teacher. When that was done for the six very different schools, the degree of common practice was striking. In all the schools, the portion of the full timetable actually spent teaching was between 84 per cent and 85 per cent.

That degree of common practice is probably historical and rooted in useable teaching units of time. Nevertheless, the salary budget for each of the six schools purchases a maximum 85 per cent of actual teaching per member of staff.

For the head and governors seeking some new margin of capacity to balance the school books, reexamining the baseline for contact time may seem the least unpalatable option available. Yet, increasing contact time can jeopardise classroom quality all too easily; the common contact baseline has been established for sound educational reasons.

But another area is worth examining. Some teachers - possibly half the staff - may receive additional remission from the contact load carried by their colleagues. But for how many periods in the timetable, and why? Answering that question involves grappling with some key dilemmas.

One concerns the relationship between payment and the task. If a teacher is promoted above the common pay scale, to what extent is the increased salary sufficient reward and incentive? If the school creates a head of department post paying four responsibility points, must the holder be given some further remission from teaching? Should post holders use some or all of their existing non-teaching allocation for their new paid responsibilities?

The marking and preparation displaced would be done after school or later, the teacher being motivated by their increased salary. That way, goes the argument, good practitioners would not be lost to the classroom.

In most schools different rungs of the promotion ladder have become associated with additional non-teaching contact time. Within staffrooms, such patterns have become a form of differential and status.

Yet why do responsibility post holders do less teaching than non-promoted teachers? It can only be because the school believes that the posts involve key activities which can only be undertaken in the teaching day. Therefore, despite the increased salary, it becomes necessary to give the post holder more freed time. Time, presumably, to hold discussions with colleagues, to prepare materials and schemes, to observe teaching and learning, to oversee resources, to liaise with outside agencies.

How do heads discriminate in practice between activities which contribute to the teaching and learning process, and those which become a form of managerialism and ends in themselves?

A comparison of our six 11-18 comprehensives shows that none give holders of a single responsibility point more non-contact time than unpromoted teachers. However, at two points and above patterns between the schools begin to diverge widely.

Almost all the schools have set their head of year posts at the three responsibility point level. All have recognised a need to allocate such roles less teaching than non pastoral post holders at the same level, on the grounds that pastoral capacity must be available during the teaching day to deal with the time-consuming and unpredictable nature of pastoral work. Heads in the sample comprehensives commonly accept that year heads need to be available off timetable in a way not necessary for other middle managers like heads of geography and history.

But there is significant variation in the level of remission actually granted to pastoral staff, ranging from 8 per cent to 20 per cent, and possibly also including protection from cover and linkage to Key Stage co-ordinators at five points.

At four responsibility points the six schools have heads of large teaching subjects (English, maths, science) with levels of teaching remission close to or identical to that for three point holders; differences appears to be as much associated with status as with the operational need for greater availability from the classroom.

Overall, however, there is less consensus than about pastoral posts. As a result, a senior head of department is allocated an additional non-contact figure of 5.7 per cent in school A and 12 per cent in school E for leading apparently similar levels of staff and resources.

It is, however, at the level of the senior management team (SMT) that levels of remission from the classroom really begin to become significant. Indeed, for any head, the level of classroom contact reasonably expected of senior teachers and deputies can be very difficult to resolve. There needs to be sufficient contact, it is argued, for such post holders to keep credibility with their colleagues and to keep close to what is after all the fundamental task of the school: teaching pupils. At the same time, the operational needs of a large organisation, and provision of the necessary problem-solving capacity, let alone the new responsibilities which exist post-local management of schools, seem to justify a generous level of remission from teaching.

Among deputy heads in the sample schools, the total level of additional remission from teaching averages 46 per cent, but with some contrasts school by school: between schools B and F (40 per cent) and C (55 per cent), for instance. Among five point post holders there are also variations worth examining. For example, between 40 per cent of extra remission from teaching in school C to only 15 per cent in school D.

The more senior a post of responsibility, the more expensive becomes the unit cost of the work done. An hour of a deputy head’s time in a Group 6 school will cost at least twice as much as the same time spent by a teacher on point 7 of the common pay scale.

So will the quality of tasks undertaken by SMT members in their increased non-contact time reflect this? Not necessarily, as Charles Handy pointed out when observing a deputy head putting out the chairs for assembly in his 1984 analysis of schools as organisations.

There is considerable evidence that much SMT work can in practice be clerical and routine, of the kind which could be undertaken at a much lower level of salary by support staff. There is even recent research indicating that the post-LMS spread of management information systems has led to senior staff spending more time typing their own memos and letters, rather than using the expensive hardware for higher-level strategic activities.

If deputy heads are expensive, then the time of heads costs even more. How should they allocate it as their non-contact time will be the highest on the staff?

One key issue for many will be how much to teach, given the powerful sense of being headteacher which still exists in the profession. Many heads still believe that they must maintain a personal teaching commitment, in order to keep both credibility with their staff and in touch with the classroom.

Studies of new heads show this can be a key issue. Other research shows that heads who teach like the sense of getting back to why they came into the profession and the respite from the unexpected, so that many insist on not being disturbed whilst teaching.

A class taught by a head is costing three times as much per period as the same pupils being taught by a mid-scale teacher. There is some evidence that the less someone teaches in a week, the less effective the teaching. This is not least because teaching becomes an interruption to other activities and is no longer the key focus.

If the head wants to keep in touch with the school’s classrooms, which is likely to be most effective: taking a top set to maintain his academic credentials; the bottom set to show that she can still hack it with the best of them, or regularly taking cover classes at random?

Among the six comprehensives, the size of the head’s teaching load varied significantly from 30 per cent of the week to 5 per cent. As the pressure mounts on schools to teach more pupils without increasing the number of staff, the issue of the contact baseline will be thrown into sharper focus unless class sizes are to rise still further.

The total amount of remission from teaching for holders of promoted posts up to and including the head is important to recognise and review, because it represents the staffing cost of managing the school as a whole. Such “management” costs tend to be historical in origin, and can incrementally increase without any assessment of the gain for pupils in reducing the teaching load of often the most effective teachers in the school.

The complex relationship between increased salary and more non-contact time can all to easily become fudged into a formula which links promotion with an entitlement to additional free periods for most post holders.

At times it is worth thinking the unthinkable. In secondary schools as a whole, the contact ratio for classroom teachers averages 0.75 nationally. In other words, all the staff appear to teach on average for three quarters of the week - the equivalent of four teachers in the school appears to average no teaching at all.

Heads and governors should wonder what would be the effect of all the teaching staff, including the head, carrying the same teaching load so that the “management” activities beyond the common pay scale were displaced outside the teaching day? Doing such a calculation is relatively straightforward. For a Group 6 comprehensive of l05 teachers, 48 holders of promoted responsibility point posts, including me, have a total of 359 periods of additional non-contact time every timetable cycle.

If none of that remission were available, the implication for overall teaching load is clear. The equivalent of six more full-time teachers would be available within the timetable. As a result, the average teaching load of all staff currently 84. Two per cent or 59 out of 70 periods, would fall to 80.3 per cent - that is, to 56 periods - three lessons less every l0 days.

To some hard-pressed classroom teachers, three more non-contact periods per fortnight may seem an attractive outcome. It may seem considerably less so to the year heads and senior post holders who already see themselves as working excessive hours and who would then face all their “management” activities taking place outside the teaching day.

Nevertheless, as class sizes and teaching loads come under pressure, time out of the classroom needs to be justified in commonly understood terms. “Management” activities, whether they involve pupils, staff or outside agencies, must be kept under review, so that their time cost is acknowledged and seen to contribute clearly to the overall quality of the educational process.

* Kevin McAleese is head of Harrogate Grammar School in North Yorkshire