That pupil numbers are rising is no secret: by 2026, we will need to find space for an extra 534,000 students in secondary schools (see bit.ly/Projections17).

What seems less well known is that, if current trends are any indication, about 11 per cent (58,700) of those extra pupils will have SEND. A further 1.7 per cent (9,100) are likely to have needs that are complex enough to qualify for an education, health and care plan. Taken together, that’s the equivalent of an additional class of 20 pupils for each secondary school in England.

If you think the solution, in part or in whole, is to increase capacity in special schools, consider this: the Department for Education projects that the population in state-funded special schools will also increase, by 14 per cent. That’s an additional 13,000 pupils by 2026. Just to accommodate them, we would need to build about 130 new special schools - in effect, replacing almost all of the 140 state-funded special schools that have been closed since 2001.

According to the DfE, 61 new special schools are in the pipeline. Just over half of the 56 free schools in this total will be specifically for autism spectrum disorder, and only 13 schools overall are planned in the Midlands and the North. It’s unclear whether the DfE has done the capacity mapping to ensure we open the right schools in the right locations for the right children. While Baroness Warnock may have welcomed former education secretary Michael Gove’s pledge to widen school choice, the opportunity to link this drive with the 2014 SEND reforms to create a coordinated national strategy for SEND was completely missed.

What does this all mean? As it stands, it means we cannot assume special schools will meet the demand predicted over the next eight years, so mainstream schools will be required to play a key role in local approaches to educating pupils with SEND.

To be clear, I am not calling this a crisis - but it is a challenge. And we need to start acting on it.

Firstly, there’s the hardware: schools. Part of the solution could be to include alternative resource provisions (ARPs) in the plans for expanding the mainstream secondary sector. These small specialist units, co-located with mainstream schools, have the potential to create much-needed capacity and provide opportunities for pupils with SEND to be taught in mainstream lessons alongside their peers. Research on the overall effectiveness of ARPs, however, is thin. An urgent review could provide the evidence on which to base proposals for expansion.

A few more inclusively minded mainstream schools, such as the one Nicole Dempsey works in, here and there, is simply not going to cut it. We require sector-wide commitment and action. Like it or not, the accountability system drives behaviour.

So, secondly, as Adam Boddison suggests, we could use it to incentivise better behaviour. One option would be to make SEND a limiting judgement for inspection: a school’s overall grade would not be allowed to exceed its grade relating to SEND provision and outcomes. That could put an end to sharp practices such as off-rolling and off-loading; no more families being encouraged to consider the school down the road because “it is better at SEND”.

Fundamentally, inclusion is about what headteacher Vic Goddard calls “botheredness”, so it would perhaps be better to reward and encourage positive actions. We could incentivise individuals by hardwiring excellence for SEND into performance management and promotion systems, developing career progression for teachers and leaders that is contingent on evidencing practice that improves experiences and outcomes for pupils with SEND.

Thirdly, we need to acknowledge those preoccupations with school organisation that may be holding us back. Research shows that pupils with SEND tend to be taught together in small “low ability” classes. The evidence is clear that setting and streaming is associated with lower-quality teaching and learning experiences, as well as poorer outcomes for pupils in these classes. As the Warnock committee found, language matters. Inclusion is about valuing youngsters, so, 40 years from now, labels such as “low ability” and “bottom set” should have become as redundant as “backward” and “retarded”.

The overreliance on teaching assistants to teach pupils with SEND has also become problematic, as Baroness Warnock and Klaus Wedell have highlighted. But this isn’t because TAs aren’t doing a good enough job. It’s an indication that the model of inclusion we have drifted towards over recent decades reflects a failure to fully address long-standing questions about how pupils with SEND are taught in mainstream settings.

There is much we can do to improve how we deploy TAs, and offer the flexible approaches to grouping and lessons Wedell recommends. But the effects will be limited without significant improvements to teaching. The Warnock inquiry concluded that increasing teachers’ knowledge of SEND was essential, and while SEND now has a higher profile in initial teacher training than it did 40 years ago, critics continue to point to its status as a Cinderella subject. The committee’s recommendation of efficiently integrating a “special education element” into ITT, rather than having standalone content that might overload the timetable, remains pertinent.

So, the fourth and perhaps most important area for action is the improvement of teachers’ confidence and competence when teaching for SEND. And not just as part of ITT, but habitually, via opportunities for professional reflection and development. Matt McArthur explains how this could be achieved in special schools. With better collaboration across the sector, we might realise Simon Knight’s vision for a more connected and collective educational endeavour within our communities.

Contributors to this Tes special have rightly emphasised the social and moral case for inclusion. Compelling though it is, we should perhaps lose our shyness about advancing the economic case. There are significant long-term economic and social costs involved in failing children and young people with SEND, as revealed in the correlations between SEND and exclusion, low attainment, being NEET, and youth crime. Early, sustained intervention not only saves money and lives but also enriches society and the national economy.

So, where now? Perhaps the next phase of this debate starts, as Nancy Gedge suggests, not by gaining consensus on what inclusion is, but by agreeing that it is important - for everyone.

Rob Webster is a researcher at the Centre for Inclusive Education, UCL Institute of Education