The new government has adopted an interventionist, zero-tolerance approach to failing schools, although research has provided few insights into failure and no one really knows how to solve the problem of poor performance at an institutional level.

Current policy is based on a fundamental misunderstanding derived from the school improvement literature. Successful schools have been studied exhaustively and their characteristics are well known. They are orderly, purposeful and well led. Visionary heads, who are firm but fair, involve governors, staff and students in effective learning partnerships.

Less successful comprehensive schools haven’t been studied much, so their problems are normally defined by the absence of “key characteristics” of successful schools. They are less orderly, less well-led, less fair, less firm. Their outcomes, attendance, test results and student behaviour are less good.

Governments have welcomed the clear policy implications. “Good practice” must be spread more widely. High homework and literacy rates are associated with good GCSE results, so every school should be encouraged to increase homework and teach basic skills. If improvement does not follow the script, less visionary, less purposeful folk should be sacked as quickly as possible. Provided their hearts are in it, all schools can succeed.

This analysis is almost wholly mistaken and offers a tautological recipe for improvement which invites leopards to change their spots without asking why they have spots in the first place.

If literacy and homework were simple, context-free elements in the educational arsenal, why aren’t they pursued with the same vigour and success in every school in the land? Poorly performing schools are seldom negative images of their successful counterparts, or St Trinian’s-like versions of the real thing. Although an occasional bad driver wrecks a school through ignorance, the pattern of accident blackspots suggests that human error is a small part of the problem or the solution.

Weak schools of the type identified by the Office for Standards in Education should instead be recognised as different in kind from the league-table elite. High-achieving models may be inappropriate and unhelpful for the less successful. Strenuous efforts should be made to analyse the distinctive features of under-performing schools, and to develop strategies for improvement that do not depend on heroic headteachers producing miracles and earning knighthoods.

The impatient “name, blame and shame” policy assumes that ministers are tackling human failure and rejects the possibility that low performance may be a successful adaptation to particular environments.

Without a better understanding of what is wrong, David Blunkett risks blaming passive smokers for their lung cancer and famine victims for the shortage of food. Greater knowledge should lead to improved treatment; short cuts are unlikely to produce significant long-term benefits.

Schools in the “special measures” category share common characteristics which are well-known but insufficiently investigated. They include: * High uptake of free school meals. No school with a below-average uptake of free dinners is in special measures; * Abnormally low GCSE scores; * Close proximity to low-status housing; * High local unemployment, with local industry and population in decline; * Attendance rates well below average; * Above-average number of statements of special need; * Above-average number of children with English as a second language; * A long history of disadvantage.

Although apparently remote from the classroom, these influences shape aspirations and attitudes in almost every “special measures” school. Social problems permeate the classroom, driving teachers into defensive postures. Youngsters lose concentration easily and challenge the less capable staff. Lessons have to be especially well organised to ensure that everyone stays on task.

Teachers are acutely aware of their students’ difficulties and worry about the unsuitable academic diet promoted by the national curriculum and GCSEs. Yet some schools thrive despite apparently adverse circumstances, so the critical task for research is to identify characteristics specific to “special measures” schools and to examine their interaction with a problematic environment.

A number of obvious areas deserve detailed investigation: Patterns of headship. Do schools in disadvantaged areas recruit heads as easily as elsewhere? What are the particular issues associated with schools at the bottom of the pile? How are energy and morale sustained? What is the career profile of heads who have worked in unsuccessful schools? How do heads counter negative factors in the locality?

Patterns of recruitment. Do schools in areas of social stress experience more than their share of “hard-to-fill” vacancies? Is there an above-average number of temporary or supply teachers? Are particular types of teacher recruited? Is staff turnover more rapid than elsewhere?

Patterns of health. What is the incidence of sickness and absence among students and staff? To what do they attribute their absence? Are patterns consistent across time? How do they compare with other areas?

Patterns of intake. How has the school roll varied in recent years? How has cognitive data (such as National Foundation for Educational Research test results, reading scores) about students varied over time? How do changes relate to the local scene? Does a critical number of above-average children make a difference to overall outcomes?

Patterns of teaching. What types of teaching achieve the best results in under-performing schools? Does the concept of “very good” teaching vary By context? Is the comparison with national averages and expectations relevant?

An enquiry into these issues should be a high priority for the Government’s standards and effectiveness unit. Until we understand the dynamics of poor performance there is little point in shouting at individual teachers and schools as if they had deliberately engineered a humiliating experience for themselves. When similar institutions fail in common circumstances, science seeks the cause, not the villain.

Bernard Barker is principal of Rowley Fields Community College, Leicester