I am writing this brief note because I cannot attend in person.

What this is about is giving a substantially greater number of children better opportunities to develop their talents and to take fuller control of their lives as adults. We cannot know if we have succeeded until these children are fully adult. Therefore quick fixes or looking for solutions over one or two Parliaments cannot work. As with most long haul questions (probably most of the issues which matter to us citizens), a political climate in which coalitions are normal will have a much better chance of delivering substantial improvements.

Who are these children? Many are from poor, single-parent families; numbers have parents with some degree of illness or handicap; but the best single predictor of which children fall into the group is poor education of both parents. All that has been on record for 50 years.

That said, there are a number of measures that have been demonstarted, at least in some degree, as likely to work:

1. Mixed neighbourhoods, so that these children live among and are in classes with kids whose parents are reasonably well educated. There is some suggestion in the relevant research that the it is when the proportion of kids with with one or both parents having graduate education exceeds 20%, the improvement in the educational performance of the rest becomes noticeable.

2. Educationally enriched nursery classes - Sure Start, Head Start or what you will - deliver improvement which shows up later in the course of education.

3. Parental choice of school has some favourable effect provided parents have sufficient information and options for choice to be real.

4. Payments to parents conditional on demonstrated school attendence produces higher attendance in this group.

5. Late in the school career, when earning is a real alternative, cash rewards related to educational attainment may work.

6. Raising teachers' expectations and opening pupils' horizons delivers improvement.

So far as my knowledge goes, that is the full list of measures backed by evidence. There are other detailed policies which I guess might work if properly and carefully implemented - extra money following disadvantged pupils is one - but they all need testing and development before introducing them on a large scale.
If we throw untested "solutions" at the problem, we are as likely to perpetuate it as to mitigate it.

Among tested policies, we need to experiment to see which most reinforce the others. Flexibility in how the measures are deployed locally, and strong means to enable lessons learned in one place to be adopted and adapted in other areas will both be essential.

David Heigham