Reforms over the past 20 years have generally embodied the assumption that we have workable solutions to some of the biggest shortfalls of our education system. The national curriculum and Ofsted were invented to address patchy standards; the literacy and numeracy hours to rectify underachievement in key stage 2 Sats; Sure Start to help those missing out on early childhood learning.

However, for perhaps the most fundamental shortfall of all the yawning gap in achievement between children of advantaged and disadvantaged social backgrounds there are no neat solutions. The push on standards and assorted initiatives to help disadvantaged children have yet to alter the basic architecture of this gap. Being poor means on average that a child is nearly a year behind better-off peers when entering primary school, rising to two years by age 14 and translating to a grade and a half at GCSE.

Those who would alter this may understandably feel demoralised by evidence showing the high impact of home background relative to what goes on at school. Whatever governments say and do about help with parenting, there are limits to how much the state or society can change what goes on in the home. Yet to conclude that nothing can be done to change the social gap in educational outcomes would be a huge mistake and a tragedy.

International studies show that while no country has abolished this gap, some have succeeded in containing it far more than others with the UK at the unfavourable end of this spectrum. And the Joseph Rowntree Foundation research published today shows that the gap is not caused just by what happens at home, but is fuelled by important socio-economic differences in the experiences of children at school and in their communities.

Primary school children from better or worse-off backgrounds differ not so much in the importance they attach to education (both recognise that it plays a key role in their life chances), but in how confident they are that they can benefit from school.

Children from better-off backgrounds enjoy rich experiences of learning, including out-of-school activities in which they see from an early age that learning can be a productive collaboration between adults and children. Worse off children, who have missed out on such activities, often view school as a hostile and coercive environment. This is not helped by the knock-on effects of pressures on schools in more deprived communities and on teachers in these schools. One of the most disturbing points in the Rowntree research is that children in deprived neighbourhoods often reported being shouted at, or “got at”, by their teachers.

New measures to address these inequalities are needed but require careful design. Extended schools could potentially provide poorer children with a wider range of learning experiences outside the classroom, but only if they can avoid reproducing negative emotions that some children have towards school. More systematic channelling of resources into schools with deprived pupils might start to improve their experiences in class, but not unless they are linked to effective changes in teaching and learning practices.

Some of the authors involved in the Rowntree research ask whether social mobility will ever be improved without a “socially critical” approach, fundamentally altering power relationships within schools, which tend to reproduce those in society.

Over the next few months, the Joseph Rowntree Foundation will be promoting debate about how to design policies and practices that reduce the education penalty for being poor. Finding effective solutions will not be easy. Yet the latest research shows that many aspects of children’s early experiences within schools could be improved.

Social background has a huge influence, but a child’s fate is not sealed by the profile of the family into which he or she is born.

* “Experiences of Poverty and Educational Disadvantage”, a round-up of findings from Joseph Rowntree Foundation research, by Donald Hirsch, can be found at www.jrf.org.uk