Will building more homes help to reduce housing costs?

As part of the Foundation’s ongoing housing work, leading economist and Intergenerational Commission member Kate Barker and Housing market analyst Neal Hudson write about the impact that boosting housing supply could have on prices and wider housing costs.

Since the mid-2000s the dominant narrative about housing in the UK has been around a shortage of supply. In its simplest form, this argument runs: the ONS projects that around 210,000 households will form in England each year. In the ten years to 2016, housebuilding statistics show an average of around 130,000 new homes were built annually. That means we have not met the housing needs of at least 800,000 households over the last decade. As a result, we now need to raise housebuilding to at least 275,000 annually to meet new demand and deal with the backlog.

This orthodoxy has recently been challenged in a series of blogs by Ian Mulheirn who asserts there is, in fact, no undersupply of homes in England today. He puts forward two lines of argument. First, that the rising number of vacant homes in England over the period indicates there is no lack of housing, confirmed by the fact that a lower number of households have formed than the ONS has projected. And second, that with the exception of London, rents and the ongoing costs of home ownership have been falling in real terms – another signal that there is no shortage of homes.

In truth, neither of these accounts is entirely convincing. The story that ascribes all the problems in the housing market to a serious lack of supply doesn’t take advantage of the fuller count of data on net additional dwellings which shows that new annual supply has averaged just over 166,000 in England over the past ten years (Chart 1). Nor does it acknowledge the possibility of changes in preferences – a rising number of students and migrants in the population over the decade means it is reasonable to assume that in aggregate, the desire to form households has changed too. And effective demand will have been subdued by the financial crisis and subsequent weak income growth.

But the argument which says there is no shortage of supply has its own limitations. To begin, as the chart above suggests, it is only when we have census data that we can accurately assess new supply. It also avoids a point about effective supply. There are a number of reasons to think an increasing number of dwellings are not part of the effective housing supply than in the past. Property is more likely to be held simply as a store of value today (no surprise that the highest concentrations of household spaces with no usual residents are found in central London). The number of second homes in England rose by 100,000 between 2008-09 and 2013-14, at least in part because of the increasing proportion of people working away during the week.  And the rise of platforms like Airbnb mean that a growing number of properties are rented out only for the very short term (there are now nearly 50,000 listings on Airbnb in London, for example, 51 per cent of which are entire properties).

The census is also our best data source for the number of households who form over time, making the gap between the household count and available housing stock the difference between two large numbers – both of which are increasingly uncertain. But the really fundamental point is that households can only form if there is a household space available for them to move into, and they can afford to do so. Young people have seen a fall in their wages relative to older age groups in recent years, and are further disadvantaged in the housing market by the increase in house prices caused by lower real interest rates. Any rise in house prices relative to incomes requires a higher loan-to-value (LTV) mortgage or a larger deposit relative to incomes (Chart 2). But at the same time, there is more caution about lending into a market where values might fall so high LTV mortgages have become less available.

This combination of larger deposits and lower earnings has trapped many young people in the private rental sector. While rents may not have outpaced prices over the period they have outpaced earnings (Chart 3). This suggests that supply is after all an issue – given these lower earnings the level of rents might have been expected also to be lower over time. 

Economic models for house price determination all conclude that rising incomes push up demand (certainly for space) in addition to the rising population.  When incomes rise, houses will need to be supplied at above the rate of (desired) household formation if house prices are to be kept rising at a rate below income growth.  We have done the opposite in England, resulting in average household size ceasing to fall, as fewer than expected elderly single occupiers are outweighed by more than expected households containing young adults living at home and multiple families.

Whether policy should seek to raise supply beyond the latest household projections, and by how much, depends on the outcomes we want in the housing market. To continue at around the 200,000-220,000 of annual new supply, roughly the present situation, would be to continue with many young people at home or sharing in the private rented sector, and many living in overcrowded conditions (which, according to the bedroom standard, has increased in the social and private rented sectors since the early 2000s to around 7 per cent and 5 per cent respectively today).  The rising number of homeless people also suggests a shortage of social rented stock.

However, increasing supply alone will not fully tackle the issue of declining rates of home-ownership.  Some of the shift into private renting may be by choice – although given the relatively insecure nature of the English rental market for tenants, it is hard to believe that since 2003 the extra one million families with children in private renting are all there because that’s what they want. As explained above, younger people are kept out of home-ownership by changes in the mortgage market and lower relative incomes.  We shouldn’t blame a ‘broken housing market’ for factors that are beyond that market, and that won’t be affected by increasing housing supply.  But increasingly supply would reduce these pressures for the young over time (could take at least a decade); homelessness and over-crowding could both be eased; rents might fall a little.  These would all be welcomed by many hard-pressed households.

There are of course wider problems about the distribution of housing beyond the narrow topic of this blog.  It’s partly because of the political intractability of those issues that we need to keep the new build rate up, in order to ensure those at the sharpest end of these wider problems can be housed decently in the future.