School improvement programmes have tried to make schools better, without aiming to make them absolutely effective. Now a group of schools in Avon have tired of “more of the same” and are attempting to generate failure-free schooling.

They have joined the Highly Reliable Schools project which aims to make schools not just relatively better but absolutely good: it seeks to eradicate failure by ensuring that all children acquire at least basic learning skills. It is based upon insights into High Reliability Organisations, like air traffic control rooms, that are not allowed to fail.

This radically-different strategy has been proposed because global economic pressures will confine the old industrial economies to pauper status if they cannot generate greater levels of skilled workers. Thailand will exceed Britain in its total Gross National Product within 15 years on current projections, and many other Asian and Pacific Rim societies appear to have what it takes to advance rapidly in prosperity.

Britain used to make its living by developing original, valid ideas and then working on them inefficiently. The information revolution means that these original ideas can now reach other countries within a millisecond. There is no future in a society having originality without reliability and efficiency; and no future in a society having an uneducated “trailing edge” of the size we have in Britain.

Much of what has passed for “school improvement” or “school development” in Britain and the United States has just increased the gap between the leading edge of schools, and the trailing edge. We all enjoy pushing up the ceiling of competent practice - raising the floor is less popular.

We believe we “know” that some practices actually work, yet this knowledge is not reliably spread. However good our ideas about education, the lack of reliability in how they are taken up restricts their validity. It is time we looked at new ways of achieving greater effectiveness. There have been useful insights from other countries, so why not model our practice upon those that appear to have virtually eradicated failure to acquire basic skills, like Taiwan?

Recent studies of school effectiveness have shown that the variation in quality within schools is much greater than the variation between them. This means that all schools have some practice that is relatively good. The problem is that the good practice is not identified and spread universally.

The HRS project aims to build on our knowledge of what reliable organisations do when they are not allowed to fail, and then apply that to educational evaluation, and school effectiveness and improvement to create a distinctly new type of organisation that will aim systematically to eradicate failure.

The characteristics of these highly reliable organisations are:

* they have a limited range of goals which require total success. For example, the air traffic controller’s job is to land the aeroplane, not to relate socially to the pilot;

* they recruit proactively and train extensively, on a pre-service and in-service basis;

* they have formalised, logical decision-making;

* they include measures to identify flaws and generate changes, for example, the simulations which test human and physical components in the nuclear power industry;

* they pay considerable attention to evaluating their performance;

* they are alert to lapses and they pay attention to detail to prevent any minor error cascading into major system failure;

* they are highly co-ordinated and interdependent;

* they are, crucially, data-rich organisations which continuously monitor how they function in order to improve their decision-making.

HROs exist when there is a public perception that failure to do something successfully - like land an aeroplane or operate a nuclear facility - would generate costs that are too great for society to bear. Such a belief is beginning to be applied to schooling in Britain

Our HRS schools are now generating very ambitious targets in a few areas, such as academic performance and attendance rates. They have tested new pupils’ intelligence and reading, to see where there is talent that can be developed. They are tapping evidence about where best practice lies, so they can compare themselves with it. They will be devising new strategies with their intakes of pupils to improve performance and avoid a trailing edge of poorly-performing pupils.

They are using focus groups of parents and pupils to report their views on school practices. And they are beginning to use performance data to trace the educational careers of children who have over-achieved and under achieved.

In three annual cycles, we propose to:

* in year one, use in-service days to bring to the schools the world’s greatest knowledge in school and teacher effectiveness, and school improvement;

* in year two, use the information systems to help schools compare themselves with their own best departments and practice and against best practice elsewhere;

* in year three, and beyond, move trailing edge practice in schools ever closer to leading edge practice.

The term “Highly Reliable School” and the use of industrial models will cause concern for many practitioners of our generation who see education as being mostly about transmitting values, not skills. They have been reluctant to take up the detail implied in the HRS model, preferring instead a broad-brush approach.

But children experience their education through detail, not the broad brush of school improvement programmes and development planning. If the education system cannot provide education for defined skill outcomes, for a very high proportion of children, generated by processes that are correct in their detail, the political system will continue to wreak its own particular kind of havoc in our schools.

The Highly Reliable School aims, simply, to transform education by generating schools which can “think”, based on high quality performance data, and which can “act” upon a knowledge of what constitutes their own and the world’s best practice.

* David Reynolds is Professorof Education at the University of Newcastle upon Tyne. Sam Stringfield is Principal Research Scientist at John Hopkins University, Baltimore, USA