A panoramic photograph of Malham Cove in North Yorkshire.

England has far too few empty homes. We should build more.

Tom Forth, .

When people in England get on a bus or a train, they are happy to see a seat empty.

When people in England go to the shop to buy milk, they are happy to see an unsold bottle rather than an empty shelf.

But when it comes to homes, many people in England think about things differently. Empty homes, they say, are a bad thing, to be reduced by policy, and to be filled before any new homes should be permitted.

Largely as a result of this, England has the lowest rate of empty homes in the developed world (OECD).

England has far fewer empty homes as a percentage of housing stock than any other developed nation. Source: OECD.

If you want to explore how the vacant home rate varies by place in England then Open Innovations have just published a new tool to help you do that.

Democracies choose their outcomes.

Government policies reflect our public preferences. Using precise measurements, estimates, and models the UK government constructs detailed mathematical formulas that predict how many households we think will live in each local authority of England well into the future. Then we strictly control the allowed supply of housing to match the predicted number of households.

The rate of long-term empty homes varies across England. Over 2% of homes (1 in 50) is vacant in Blackpool compared to fewer than 0.5% of homes (1 in 200) in Manchester. But in all places the rate is unusually low compared to other countries. This is exactly what we’d expect from a process that tries to precisely match supply and demand everywhere.

Lessons from abroad.

Across England, 0.9% of homes are long-term vacant. The equivalent number for France is 8.2%.

In Greater London 0.7% of homes are long-term vacant. The equivalent number for Greater Paris is 6.5%.

The only part of France that comes close to the low levels of empty homes we experience right across England is Corsica with a rate of 3.1%, largely due to a high proportion of second homes. The people of Cornwall can explain how that works much better than I can.

In France the goal of housing policy is different to in England. There are far too many details to cover adequately here, but one strong philosophical difference is that the French trust markets more.

Instead of complex government models to calculate household count projections, and then an attempt to micro-manage housing permissions and supply to meet those projections exactly, the French are happy to target an increase in empty homes. Indeed the current plans for housing in Greater Paris think that the vacancy rate of 6.5% is too low and they hope to build enough homes to increase it to 8% by 2035.

To achieve this, local, regional, and national governments right across France continue to grant permissions, make investments, and assist with building about 50% more homes per year than we manage in England.

So what?

There are many factors that influence the price of buying a home and the cost of renting one. Housing supply is just one, but it is one that our government can influence quite cheaply; it is largely the private sector using private money that builds homes once they have been given planning permission.

When we build more homes we create more empty homes. And because there are more empty homes, prospective homeowners and tenants have more choice, more chance to shop around, and more chance to pay less for what they’re looking for. Many choose to buy or rent a larger home or a higher quality home instead of paying less for housing, but without a good method for comparing housing quality across places this is hard to track.

What we do know is that French of all ages and incomes pay less for housing than people in England, and in part it’s because there are more empty homes for them to choose to live in.

There are other advantages to having a slightly large stock of empty homes too. It makes moving house far easier. Not only do you have more choice of properties, but transaction chains are shorter, cheaper to manage, and less likely to collapse. You are far more likely to be able to choose the time and date that you move. Empty homes are also the easiest homes to renovate making it easier to do up a property or make it more energy efficient quickly and cheaply without having to work around the occupants.

We could do the same in England as we observe in France. Instead of using huge amounts of data and complex models to try and precisely match the supply of housing to the number of households, we could change our thinking to allow sufficient building to reach a healthy level of vacant properties. 8% seems like a decent place to start, though 10% might be even better. Whatever the detail, the number should be a lot more than it is today if we want to lead better lives in better homes.

People in the UK (and thus in England where around 80% of UK households live) pay too much rent, in part because we have too few empty homes. Source: OECD.

What are all your sources for this data and are you sure it’s right?

I’m not sure that this is right. But I’m pretty sure and I’ve spoken to dozens of people and tried hard to test this data against alternatives. I am always keen to learn more and talk to more people and improve. A simplified version of my understanding as of today is below.

The thing most people suggest I read to understand that I’m wrong is Tackling the UK housing crisis: is supply the answer? by Ian Mulheirn. I have read this and much of his further writing on the topic. The core method is to use the same dwelling estimates as I do, Table 125: dwelling stock estimates by local authority district, and estimates of the number of households from the Annual Population Survey. Estimates of the number of households from the Labour Force Survey can also be used for this purpose. No matter which source of household estimates is used the estimate has high uncertainty at below the scale of all England. An estimate of “surplus dwellings” is calculated by subtracting the number of households from the number of dwellings.

This method shows an increase in the number of “surplus dwellings” in England from approximately 800,000 (4% of dwellings) in 2002 to approximately 1,200,000 (5% of dwellings) in 2019 and is used by many to argue that the counts of vacant homes reported by local authorities and collated in Table 615: vacant dwellings by local authority district: England, from 2004 are incorrect to show decreasing numbers of vacant homes.

There is some reason to think that local authority reported numbers of vacant homes could be underestimates. While in the past many local councils offered council tax reductions for vacant properties they increasingly use their powers to charge full council tax or council tax at the full rate plus a supplement to long-term vacant properties. Where in the past the homeowners had an incentive to report their home as vacant, even fraudulently, they are increasing incentivised to report their homes as occupied when it is vacant.

But there are also good reasons to think that local authority reported numbers of vacant homes are of high quality. The ability to charge a council tax premium on vacant properties means that councils are heavily invested in correctly identifying them. Leeds City Council has a team dedicated to the task , works with a CIC on the issue, offers advice and loans to homeowners looking to bring vacant properties back into use, and encourages the public to report empty homes. They report their own data on vacant properties in more detail than required by the UK government. They see no evidence that they are missing a significant number of vacant properties.

So what might explain the discrepancy of an approximately 1 percentage point fall in the percentage of homes that are vacant according to local government and the approximately 1 percentage point (400,000) increase in ‘surplus homes’ suggested by APS and LFS household estimates? There are two major places to look.

The first is second homes. Annex Table 1.3 of the English Housing Survey 2018 to 2019: second homes - fact sheet estimates that from 2009 to 2019 at least an additional 208,000 homes became second homes for use by friends and family in England. If the rate of growth of second homes has been constant this would explain 350,000 of the ‘missing’ homes in the period since 2002. My understanding is that second homes owned by foreign households are not included in this estimate, which would increase this number. The English Housing Survey does not consider second homes and once a correction for this is made its estimate of the number of vacant homes is of a slight decline in percentage of total dwellings in the last fifteen years.

A second source of the discrepancy is likely to be purpose-built communal accommodation, particularly importantly purpose-built student accommodation, of which the UK currently has nearly 700,000 units, with the vast majority in England. These units are counted as fractional dwellings (typically one third, but varying depending on the number of local dwellings freed up by their construction) but their occupants are not always counted as households since “ Students living in halls of residence are sampled via the private households of their parents”, an especially large problem with overseas students, whose numbers have increased in recent decades.

It's usually at this point of thinking about vacant homes that I get frustrated by the pointlessness of it all. By whichever measure we use, we have too few homes for there to be an efficient market in housing. Very specific details of survey design and small uncertainties due to sample size matter enormously to our understanding of the housing market because we operate with such small slack. We import estimates of pricing elasticity with respect to supply even though no other country has as acute a shortage. We should expect much higher elasticity at lower vacancy rates.

If we just built more homes, targeting a vacancy rate of 8% to 10% as in Greater Paris the complexity of the data would become far less important. Instead, as is increasingly common with UK government, we imagine that through more data, more dashboards, and more micro-management we can somehow deliver better outcomes for our population than a system free to respond to market signals could.

England has too few empty homes, everywhere. The shortage is greatest in cities and the South East of England, but our government-enforced matching of homes and households has made it a problem everywhere. The path to a better-housed and happier society is to step back from the data and the algorithms, and trust markets more. We should target a healthy vacancy rate of nearer 10% of homes and build the homes we need to get there.

How those homes are delivered will necessarily vary by place. In Cornwall the attractiveness of the area to second homes owners with savings well beyond what could be accumulated through local work demands a local solution. In Greater London the huge attractiveness of a world city and its strong economy means that low-paid workers are unlikely to ever be able to afford good quality housing at market rates and this demands a local solution. Other places will have similarly unique circumstances.

But in all parts of England building more homes and targeting a higher vacancy rate instead of trying to exactly meet household predictions will improve housing.


blog comments powered by Disqus