Yes, we must build more homes

Yes, we must build more homes

This is my last ever blog for Shelter, so I hope I can be forgiven for indulging in a bit of arguing-with-men-on-the-internet.

Challenging the consensus

There is no denying that there is now a well-established consensus that England has not been building enough homes for many years, so it was probably inevitable that it would get challenged. We should welcome that: challenging the received wisdom is always a worthwhile check on complacency. But the truth in this case is that the consensus is right, albeit for the wrong reasons, and the challenger is wrong, despite being right about much of the detail.

Ian Mulheirn of Oxford Economics has been writing a long series of blogs about the need for new homes in England that are well worth reading – and have inspired numerous responses both supporting and opposing his arguments. In summary, Ian argues that building more won’t make homes cheaper, the problem with housing is not supply but demand and distribution, and that we are already building more than enough homes. All of which is definitely a challenge to the consensus view that homes are very expensive, and that building more of them is essential.

Many of the pieces of data analysis that Ian deploys have been challenged by various experts. For example, James Gleeson claims that Ian is wrong to use household formation figures to show that more homes have been built than new households have formed, and that a different data source shows that the number of ‘family units’ formed each year far exceeds the new homes provided. Andrew Lainton makes a similar argument about the number of ‘concealed households’, and adds that we will always need housing supply to run ahead of household formation, not just to match it. Ian’s emphasis on net supply, rather than new building, is similarly challengeable: just as gross house building figures are distorted by demolitions, net supply figures are distorted by conversions, which increase the number of units without actually increasing the amount of housing space for people to live in.

But I don’t want to get stuck on these technical arguments, because even if Ian is wrong on some of these points, the central argument he is making here is essentially correct: house building alone will not make homes more affordable for the millions that he accepts are priced out of ownership.  According to the recently-reheated NHPAU model, a 1% increase in housing stock would only reduce house prices by 2%. As Ian  (and Peter Saunders, and Simon Wren-Lewis, and Chris Dillow) argue, demand is much more of a determinant of house prices than supply – and demand is influenced by a huge range of factors over and above the simple number of people, or households, or family units.

Right, right, wrong

So even if his analysis of the data might be contested, Ian’s first claim is true. His second claim is also sort-of true: that the problem is really one of distribution. It may well be narrowly true that there are enough homes, or rooms, for everyone in the country, as Danny Dorling has long argued, it’s just that many of them are in the wrong place or occupied by too few people, meaning other people experience a housing shortage. But this is only sort-of true, in that it really doesn’t help much. Housing is extremely hard to redistribute fairly: no-one really imagines we can force every under-occupier to give a bedroom or two to the overcrowded. And as Paul Swinney argues, low prices and empty homes in Burnley don’t help Bristol. Tom Forth puts it pithily: ‘England doesn’t have a housing market’- it has a series of different local housing markets.

This problem hints at my fundamental objection to Ian’s argument: it’s just not useful, as it doesn’t lead to any meaningful solutions.

Even if there are enough homes in theory, in practice we do need to build more – to keep up with rising population, provide accommodation for the hundreds of thousands of families struggling without decent housing, update aging housing stock and ensure that there are homes in the place where the jobs are.

Some of these problems could and should be addressed by things other than housebuilding: for example, industrial policy could seek to relocate jobs from the high-pressure housing markets of the south to places with more homes. Higher wages and more housing benefit would reduce the pressure on families struggling to pay the rent or find a home. But none of these things are easy, and there is absolutely no guarantee that they will happen if we abandon the drive to build more homes.

Ian himself agrees that we need to provide more social housing, and to tackle homelessness. In theory government could do that by buying up existing homes and converting them into social housing for those in need. In some places that’s quite probably a good strategy. But in the places that need it most it would be ruinously expensive – and wouldn’t do anything to tackle the geographic imbalances in the economy. By far the most efficient and politically realistic way to provide more social housing is simply to build more of it.

Simply put, building new homes is not the only way to solve housing problems – but pretty much all housing problems are lot easier to solve if you have plenty of homes, and new building is obviously the best way to increase and improve the stock. My real fear is that Ian’s arguments will be seized on by those who want to oppose new housebuilding (who are normally are very well housed themselves) – and that they will not replace housebuilding with robust demand-side policies, increased welfare spending, or a draconian redistribution of existing homes. As Matthew Parris says, demand-side solutions are even harder to pull off than increased supply.

Cosy consensus deserves to be challenged, especially when it slips into simplistic shibboleths. But the consensus on building more homes is still the best hope we have of solving our chronic housing problems.


Thanks to everyone who has read, shared or commented on my blogs over the last six years.