Over the year to March 2015, 156,000 new homes were started in Great Britain, of which 140,500 were in England. This is the highest number of new starts since the downturn in 2008 but it is still just over half of what we should be building. To meet housing need, we must deliver between 240,000 and 245,000 new homes in England per annum, 35,000 in Scotland and 14,000 in Wales.

The current levels of housebuilding therefore leave us with an annual shortfall of 136,000 homes across Great Britain, including a shortage of 100,000 homes in England. Unless further steps are taken to boost capacity, accumulated shortfall will rise to around one million over the course of the next decade.

CLICK ON IMAGE BELOW TO ENLARGE

^{Source: DCLG, TCPA, CCHPR, Scottish Government, Welsh Assembly}

However, our analysis shows that there is potential to deliver 205,000 new homes a year in England alone provided key barriers are eliminated. One of the major constraints is the restricted availability of development land, particularly in areas of high housing need.

While the number of planning consents is up, with permissions for 200,000 homes granted in England last year, this number not only still falls short of need but also is not necessarily located in the right places.

Building starts

We have seen an increase in new homes starts over the last two years, from 97,000 in England in 2012 to 140,500 in the year to March 2015. However, this progress could come to a halt if housebuilders are not able to replenish their pipeline of consented land, particularly in the high demand markets of Surrey, Hampshire and Berkshire, and in the East of England, including Cambridgeshire and Suffolk.

By measuring the three-year average of consented residential units against the three-year average of housing starts for each local authority, we can assess whether there is enough land coming forward for residential development, and highlight pressure points where there are insufficient sites to maintain the current state of building.

These pressure points include: Babergh, South Bucks and Worthing, where the three-year average of new homes starts is almost five times higher than the three-year average of units consented. In Merton, the rate of new build starts outstrips consents by 4:1, and in Elmbridge and Watford by 3:1.

The problem is not confined to high value and high demand areas or places constricted by the Green Belt; both Kingston upon Hull and Newcastle have a three-year average of new build starts that is over four times higher than the three-year average of new consents.

CLICK ON IMAGE BELOW TO ENLARGE

^{Source: HBF, DCLG, TCPA}

Planning consents

There is also concern that the consents that are being granted are not in the areas where there is the greatest housing need and where it is possible to build and sell the fastest.

Map 1 shows the three year average of detailed consents granted in each local authority against the housing need specified in the local development plan. Areas in red are those where the number of consents granted falls below the assessed need in an adopted local plan.

CLICK ON IMAGE BELOW TO ENLARGE

^{Source: Savills Research, Glenigan}

The problem is particularly acute in parts of the East of England and the South East where there is high demand. In Oxford, the 3-year average of detailed consents granted between 2012 and 2014 would meet 12% of housing need, in Brighton & Hove this amounted to 16% of need, and in East Cambridgeshire, only 7%.

On the flipside, the map also highlights in blue that some areas appear to have sufficient consents. In Allerdale, Flyde and Stafford, for example, the three year average of consents outstrips the objectively assessed need by over 2:1.

Local authorities with no recent figure for housing need are shown in grey. Analysis of those areas using household projections reveals a similar pattern of a shortage of consents in areas of high housing need, particularly in the local authorities neighbouring London.

The link between housebuilding starts and residential market transactions

Our analysis of private housebuilding and wider market transactions in England shows that the private sector has built to market capacity in a relationship of one new home for every 10 sales across the whole market over much of the last 25 years.

CLICK ON IMAGE BELOW TO ENLARGE

^{Source: HMRC, DCLG, Savills Research}

While the ratio is currently 1:9, demand side interventions such as Help to Buy have failed to alter this ratio substantially. If this ratio persists, private housebuilders are unlikely to deliver more than 121,000 homes in 2020.

However, if this were to shift to a relationship of one new home for every seven sales, then private sector starts could reach 150,000 in 2020. This ratio was last reached in the first quarter of 2006, but was not sustained. To permanently alter the ratio, we need a higher number of planning consents to be granted consistently in areas of high housing demand, where it is possible to sell homes most quickly.

This still leaves a shortage of 90,000 homes against the 240,000 number we need per year. It is therefore clear we need to encourage a broader range of developers with different business models including local authorities and housing associations, to deliver a wider variety of tenures including building homes for market rent (see Alternative Developers).