Thursday, 15 September 2011

Housing Crisis

Numerous reports recently have highlighted the issue of affordable housing the UK, the most recent coming vociferously from the National Housing Federation.

The most obvious indicator for a lack affordability is the rate that house prices have gone up over the past half century.

House price data from Nationwide, RPI data from ONS

This plot shows UK average house price since 1952 with data from Nationwide. The lower line shows the 1952 price adjusted for inflation (RPI). So if house prices had increased in line with other typical goods then the average price now would be around £40000. An emerging boom-bust pattern can also be seen in the plot.

Perhaps the best way to measure the affordability of housing is to present the average house price as a function of average wages.

Data from ONS
This plot shows the ratio of average earnings to average house price oscillating between 2.5 and 3.5 until the mid 1990s from where it shoots up to around 5. Despite recent falls in house prices these data show little impact on the ratio, presumably since wages have also been hit in the current economic climate. One obvious change from long term patterns was the level of mortgage lending between 2001-2008 which increased increased significantly, with mortgages of 100% loan-to-"value" and 5x earnings not uncommon. I will return to this later.

But how much of this a supply and demand issue as highlighted by the NHF? The plot below shows housing building over the past 60 years.

Data from ONS

There has clearly been much lower levels of house building since the 1970s (there were also very high levels of house building during the 1930s not shown here). This plot also shows how publicly funded house building (council housing) has declined to nothing. This means that we are now entirely reliant on privately funded house building and hence the business cycle. This is clearly a big problem, with the current economic climate, large scale public funded housing would be a strong counter cyclical force for the economy but we have lost that potential. Supply of new housing has clearly dropped significantly but, given relatively modest increases in population, this on its own cannot account for a trebling of prices in a dozen years.

Another important consideration is the evolution of the division of tenure type across the current stock of housing.

Data from ONS
This plot shows the rise of owner-occupiership in the UK during the last century and the rise of social housing until 1980. Private renting declined as more and more people had the option to leave over-crowded housing conditions and find their own home. In 1980 there was a pivotal political decision to switch to a model of private home-ownership. Since then the proportion of people in social rented accommodation has fallen and owner-occupiership peaked. The past decade has seen the first rise in the proportion of people privately renting for a very long time. It remains to be seen how far this may head back towards patterns of the last century. Another key observation from this plot is a point reached around 1970 where a majority of people are living in a home that they own. At this point there will be less political pressure for new house building, indeed there is the potential for a majority to oppose new housing since it could be seen to adversely effect a personal and costly asset. In addition, increasing house prices can now be promoted politically as a positive action since home-owners get the impression of increasing wealth without obvious effort on their part. Plot one suggests that the significant divergence of house prices from general inflation starts around 1970.

This final plot compares land values and mortgage lending over the past number of years.

 residential land value data from ONS, mortgage lending data from the Council of Mortgage Lenders.
The first observation is the spectacular rise in both land values and gross mortgage lending. Over the 12 years to 2008 residential land values increased an incredible 550%. Mortgage lending also increased dramatically doubling in just a few years and rising further before collapsing back to levels of a decade earlier. I would very much like to see more historic data for both mortgage lending and land values to extend this plot but I have not yet found it. It is interesting to note that the drop in land values, to around 60% of their peak, is significantly more than the drop seen to date in house prices (first plot). The second observation is the striking correlation between the two variables. There is a small time lag in land values behind mortgage lending. This may well be an artefact of the data recording but if accurate would demonstrate mortgage lending driving land values up, rather than mortgage lending responding to a increasing "real" value in the land. Either way this plot has "bubble" written all over it.

You can also clearly now see that the cause of the sudden drop in house building in recent years in plot 2 is down to plummeting land values. Developers who bought land between 2002 and 2008 potential now face a loss if they build. Given that we can see that we are now totally reliant on privately funded house building the prospects for an increased supply do not look good. That final plot suggests most recently land values holding above mortgage lending, potentially due to the unprecedented cut in base rates to 0.5% and the drastic drop in the number of property transactions. For them to converge again either mortgage lending would increase, or land (hence house) values drop further. Given the still significant proportion of owner occupiers who had been gifted large "paper wealth", falling house prices are not politically popular, even though in the long run the vast majority will benefit from lower housing costs. On the other hand no mortgage provider is going to back falling or over-valued assets. Equally, already heavily indebted borrowers are not going to take more debt on in the face of falling house prices and terrible job prospects. A quick check confirms that there are plenty of mortgages available at 2.5 or 3.5 times earnings with a reasonable deposit which, up until 2001, was the normal situation. So there appears to be a reality gap between house price expectations and house price values as funded by real earnings.

All this leads to a depressing stale-mate situation, with both economic stagnation and terrible prospects for young, potential home-buyers the most likely outcome.

1 comment:

  1. You put a lot of work into that! Good summary, keep up the good work.