The need for new homes is inescapable, but how many and where to put them? Gerald Kells gives his view

It seems impossible to read anything about housing these days without being told we must have three million new homes in England in the next 20 years and that this is needed because houses are unaffordable. Even though few people could say how the figure of three million is derived, it is rare for anyone to ask the question: do we really need that many houses, will houses generally become affordable if we build so many and how would it change our towns, cities and countryside if we do build them?

And they're timely questions to ask. The Regional Assembly in the West Midlands, having caved in to the Government time and again on housing numbers, has finally issued a press release warning against playing the numbers game, following the latest enforced numbers hike from 365,000 to 420,000 new dwellings.

So how many homes do we need in 20 years time? The truth is we don't know. The Government's approach is to extrapolate forward past trends and create a figure which may or may not come true. Factors such as birth and death rates, the proportion of people living alone, and immigration all vary considerably over time and are very difficult to predict.

Work we did last year showed how fairly small changes in underlying assumptions drastically change the results. Rather than take the bullish line of the Government, we believe we should build the houses we know we will need, not those we might need. After all, we will be reviewing housing need throughout the next 20 years and it is much easier to allocate more land for housing if we need to than to row back should we get it wrong.

And they'll still be a lot of house building. CPRE argues that somewhere in the region of 285,000 homes could be provided in this region in a way which supports our wider sustainability goals and provides the houses people need.

But the Government isn't just saying it's about providing numbers of homes, it's also about bringing prices down. Those of us brought up in an era where a teacher or nurse could buy their first house on three times their salary will have a lot of sympathy for young people looking at houses which cost up to ten times what they earn, stretching themselves into debt and mortgaging their futures, or worse, in a house price crash, being thrown out on their ear.

The trouble is that the driving force for high house prices is not supply, it's our willingness to pay. Even the Government's own advisers, such as Kate Barker, do not pretend that increasing the number of homes will stop house price inflation. As long as we are prepared to stretch ourselves, often on two wages, and even use up our inheritance as a deposit, as long as investors in the buy-to-let market see housing as a better bet than a savings account and as long as banks are prepared to fuel our indebtedness, house prices will remain high.

And the CPRE can back up this case. Unlike the Government's theoretical analysis of supply and price, we commissioned work to look at real areas of the country which compared both the supply of housing land and the number of houses built with price. There was no evidence that building sprees reduced prices, in some areas prices actually went up.

So the Government's claim that increasing supply will, over the longer term, impact on prices is both highly questionable and impossible to measure against other bigger and more immediate determinants. The harsh truth is that if the Government wants housing to be affordable it needs to address the real causes and not rely on a long term supply-side gamble.

We certainly do need more rented and subsidised housing within the housing we build, because for some people, market housing will never be affordable. We also need to ensure the right types and sizes of homes in areas of poor affordability. Allowing housebuilders to provide large detached dwellings in the countryside will have negligible impact on the price of rural starter homes and flats.

Some will respond to this argument by saying: it's all very well to say we might not need these houses or that building them won't bring down house prices, but surely it doesn't harm anyone to build too many. Unfortunately it does.

The Assembly's current Regional Spatial Strategy seeks to influence developers' decisions on where and what kind of houses they build. It encourages an increase in house building in urban areas while seeking to concentrate rural housing on meeting genuine local need.

To support the RSS, specific regeneration packages are being developed, for example, through the ground-breaking work of the Black Country Study.

And there's a very good reason for this approach. Our cities are losing roughly 12,000 people (net) to the rural areas every year. Because the people who leave tend to be both more aspirational and richer - this leads to social polarisation and environmental decline.

Such an approach, however, is only possible as long as planners can identify enough urban land to support housing development and can control land releases in the countryside. Increasing the housing numbers beyond a certain point inevitably leads to the release of more green field land, including, as is currently proposed, large extensions into the countryside around Coventry, Worcester, and many other towns. Suddenly our attempts to get new homes, especially up-market housing, back into our urban areas founder because the developers can prioritise greenfield sites.

And even if they do continue to build in our cities, they'll no longer be interested in those critically important higher value properties which ensure a good social mix.

Of course, the jargon is that these new developments will be part of "sustainable communities", but it is hard to see where the infrastructure, schools and hospitals will come, let alone how flood plains will be avoided. Some of the supposedly sacrosanct green belt, which is the green lung for many communities, would also have to be sacrificed. There is a nagging doubt that we will create yet more soulless housing estates, poorly connected to their local centre but with distributor roads disgorging traffic every morning onto our congested motorways.

This extra house building will consume scarce land, add to congestion, threaten wildlife habitats, increase urban sprawl and increases carbon emissions. And, having hoped to get more homes overall, we will find our housing spree has left our less desirable areas falling into disrepair and threatened with the kind of market failure that has dogged northern towns.

That is why over-providing houses is more dangerous than under-providing. And that is why CPRE argues, not that our communities should be set in aspic but that we should build the houses we genuinely need and be extremely cautious about opening up large amounts of greenfield land.

Unfortunately, as the Government's recent intervention shows, a much more simplistic ideology is driving the current top-down imposition of housing numbers which has little to do with creating communities fit for the twenty first century. If, like the Emperor's New Clothes, few dare challenge it, it does not mean the policy justification is any more substantial.

* Gerald Kells is West Midlands regional policy officer for the Campaign to Protect Rural England