What if the market is constrained by demand and not supply?
Here’s a question for you; what would be the impact of government policy increasing new supply if the market was only being held back by demand and not supply?
During the last decade, in spite of popular perception, the UK was actually delivering enough housing to meet new demand. However, since the beginning of the recession, new supply has collapsed and delivery is 33% below peak levels and housing stock is only expanding at 120k dwellings per annum while I estimate that new demand is closer to 150k households per annum (the frequently quoted 250k households per annum is unrealistic for a number of reasons).
On the face of it this would suggest a new supply shortfall of at least 30k dwellings per year but overall market turnover is 46% below peak levels and new build delivery, having compensated for a lack of overall supply during 2008/09, is now firmly inline with overall market turnover.
Meanwhile demand remains severely constrained by the inability to raise large deposits and access mortgage finance while the pricing out of first time buyers has brought house purchase chains to a juddering halt.
If we therefore assume that the supply of new build has reached a point of saturating market demand for it at existing constrained demand levels then what would be the impact of increasing supply?
Basic economics would then suggest that without corresponding policies to increase demand, any increase in supply would lead to price falls and this would eventually help to increase market activity but probably not in the manner intended. Therefore, any increase in supply would need to be counterbalanced with support for increased demand whether by subsidising mortgages for prospective owner occupiers or increasing demand for institutionally held private rented stock.
Meanwhile, as one of the few locations where demand for new build remains strong, central London developers would probably reap the rewards of any policy designed to increase supply by reducing s106 contributions.
It’s all about the under-supply, right?
Persistent under-delivery of new housing relative to demand has caused house prices to soar and priced many people out of the market, right? Wrong: housing delivery has been in-line with new household formation and forecasts for 250k new dwellings per annum are far too high.
The most common explanation you will find for the unaffordability of the UK housing market is that the demand for housing far outstrips new supply.
The official 2008 based household projections, from which forecasts for housing demand are made, puts new household formation at approximately 240k households per annum while the expansion of housing stock peaked at 207k dwellings per annum in 2007 and is currently expanding at 121k dwellings per annum. From this, the obvious conclusion reached by many in government and the market is that we were under-delivering by 33k units even at the peak of the market and we are currently building half the required amount.
However it is not that simple. The household projections probably overstate the international migration component over the medium term (even the OBR use a lower migration assumption) but the largest error comes from the conversion of the population into households. During the ten years period from 1991 to 2001, the average size of the UK household fell from 2.50 people per household to 2.41. Since then, household projections have assumed that this scenario would continue and projected the ratio to fall to 2.35 in 2011 and 2.27 in 2021. However the 2011 census has showed that this ratio has remained fixed at 2.40 and therefore the household projections have significantly overstated the level of household formation since 2001.
We actually find that the number of new households formed in England during the period 2001 to 2011 was 154k per annum compared to an increase in stock of 161k per annum which are both below the 187k per annum predicted by the household projections.
Looking forward at the increase in households between 2011 and 2021 and the frequently quoted 240k new households per annum looks highly unrealistic and a more appropriate benchmark would be using the 2010 low migration projections with a fixed household size. This forecasts housing demand of 152k households per annum over the next ten years which is very similar to the previous ten years. Under this assumption we currently have an under-supply of housing in the region of 32k units per annum.
In conclusion we do not have an existing housing shortage but a small one is possibly developing as a result of lower housing delivery. Rather than targeting massive increases in new supply across the country we need targeted delivery of units in select locations at appropriate prices.
Census 2011: household size and housing demand at a local level
The average size of households stayed static at a national level between 2001 and 2011 despite official projections previously suggesting a sharp drop.
But the national picture hides a wide divergence as seen in the map below. London had the largest increases in household size (particularly east London) and other urban areas also saw large increases (Leicester, Oxford, Nottingham, Luton and Manchester) while parts of the North East, North West and South Wales saw the biggest decreases in household size.
Although probably not the whole story (some immigrant groups may live in larger households, increases in university students sharing etc) the map gives a good idea of where affordability and a lack of housing supply may have prevented new households from forming. We will be able to investigate this further with the second release of 2011 census data beginning in November.
The map is created using my custom Microsoft Excel mapping system.
Census 2011: housing supply shortfall - regional rather than national?
The lack of new housing supply relative to demand is regularly cited as being a major driver of national house price growth and with the initial release of the 2011 census today we can look at this in more detail.
Across England the increase in new households was broadly in-line with the net change in dwelling stock (about 5,000 less dwellings than households). Combined with the fact that the number of people per household has remained constant during the period and this suggests that perhaps supply isn’t as constrained at a national level as some people believe.
However if we look at the figures for London, we can see a real divergence in the increase in stock versus households (22,000 more new households than dwellings) and an increase in the number of people per household from 2.38 in 2001 to 2.5 in 2011. This suggests a major constraint on new household formation and a simple exercise in applying the 2001 people per household ratio to the 2011 population suggests a potential supply shortfall of 193,000 units across the period.
As I’ve investigated here, a lack of supply has been much less of a driver of national house price growth during the 2000’s than previously suggested and much of the blame/credit goes to a loosening in the amount mortgage lenders were willing to lend.
However, at a regional level, there are severe imbalances and without targeted housing delivery in appropriate locations (London & the south of England) with appropriate types (family housing) at affordable prices; London will become an increasingly unaffordable/uncomfortable place to live.
Just to add to the confusion, it looks like the ONS releases are comparing the 2011 census data to the 2001 mid-year estimates rather than 2001 census but I’ve chosen to compare the two census data sets.