London’s housing non-shortage

A while ago I wrote a series of blogs setting out the evidence on what’s actually happened to UK housing supply over the past couple of decades. This blog takes a look at the regional picture.

To recap the national story, the available data show that, contrary to common perception:

  1. The number of houses has grown significantly faster than the number of households in the UK over the past 25 years;
  2. The cost of housing — distinct from the price of houses — has been falling relative to prices and earnings since at least 2005.

This suggests that whatever the problems in the UK housing market – and there are real problems: sky-high prices, a lack of social housing, or falling real incomes hitting affordability thanks to the plunge in the pound — they have not been caused by a ‘chronic undersupply of housing’. In particular, since a dearth of supply cannot explain the roughly 150% real terms house prices increase over the past 20 years, it is very unlikely that the perennial wish of housing commentators to simply ‘build more houses’ will make any meaningful dent in prices, as Kate Barker recently pointed out.

“Okay, but this can’t be true in London…”

A common response to those posts from some thoughtful people was that there might not be a national housing shortage, but there must be a shortage in some regions — most obviously London.* And it’s certainly true that London has seen extreme house price inflation recently — up by around 60% in nominal terms over the past five years. So is there evidence of a general under-supply of housing in London or the wider South East that could be driving the red-hot house price inflation there since 2013?

There are two sources of data we can use to judge this. One is DCLG’s regional housing stock data combined with the ONS’s regional household count. The chart below plots change in the number of households against change in the number of dwellings, by region, for the period 2001 to 2015. Any region sitting above the blue diagonal has added more dwellings than households over the period.

Strikingly, London and the South East lie well above the line — London’s household count has grown by just under 10% but it has added almost 12% more houses — suggesting that supply has comfortably exceeded household growth in these regions since the turn of the century.

Note: ONS households by region series (using a three-year rolling average); DCLG stock from Live Table 100

That’s how things have changed, but where does that leave the number of houses relative to the number of households? The chart below shows how the ‘surplus’ stock in London appears relatively healthy and in 2015 stood at over six percent — some 200,000 more houses than households.

Note: ONS households by region series (three-year rolling average); DCLG Live Table 100

This seems surprising, and the regional data on household numbers are inevitably less precise than the national estimates. So what about that second source, the English Housing Survey?

The EHS (and its predecessor the English Housing Conditions Survey) covers a randomly-chosen sample of dwellings around the country, identifying vacant houses as it goes. Comparing the vacant counts in 2001 and 2015 shows that the proportion of ‘surplus’ housing stock has increased substantially in almost every region. London added almost 50% to the absolute number of vacant dwellings over the period.

Note: English Housing Conditions Survey (2001); English Housing Survey (2015)

The measures of surplus, and their evolution, from these two sources are very close for England, suggesting that in 2015 there were around 1.1 million more houses than households. They don’t quite agree at regional level — at least partly just because both are drawing on samples so it would be miraculous if they did — but the direction of travel indicated by both is the same: a surplus of housing stock that’s been growing since the start of the century in most regions, including London and the South East.

The implications for rent and prices

The benign supply picture in London and the South East is also reflected in the cost of housing as captured by rent. Adjusted for inflation, rent levels fell in almost every region of the country in the decade since the ONS’s rent index began in January 2005. In London they did rise in real terms, but managed just a 4.4 percent increase over the subsequent decade — equivalent to about 0.4 percent average real annual growth.

By contrast, house price growth in London was almost ten times greater up to 2015, indicating that, whatever its drivers, a lack of places to live was not a significant factor.


None of this is to deny that there are real problems in housing: sky-high prices entrench intergenerational inequality, a lack of social housing to offer sub-market rents prices less well-off people out, and falling incomes mean housing costs are a growing burden. But these problems were not caused by, and will not be solved by, simply building more in aggregate. Only once we stop collectively barking up the wrong (housing supply) tree can we move on to thinking about the real causes of these problems — and therefore have a chance of resolving them.

— — — — — — — — — — — — — — — — — — — — — — — — — — —

*Incidentally, it’s pretty much impossible to make a case that purely regional supply shortages can explain a boom in national prices in the absence of a national supply shortage. If there’s no shortage nationally then any growing demand in London, say, must come at the expense of falling demand elsewhere leaving average prices unchanged.

One clap, two clap, three clap, forty?

By clapping more or less, you can signal to us which stories really stand out.