Councils are granting enough planning permissions – so why aren't we building housing?

The good old days. Image: Hulton Archive/Getty.

One of the central housing objectives of David Cameron’s government was to liberalise the planning system and increase the amount of land that was permissioned for residential development.

To be fair, in that respect, it didn’t do too bad a job. The number of units given planning permission in England increased from 176,209 in 2011 to 261,644 in 2015. The planning system is now yielding enough permissions to meet the roughly 250,000 new homes many housing economists think we need to keep up with household growth.

This doesn’t mean they are all in precisely the right places – although London’s permissions are running at around 50,000 a year, according to the latest figures from the Department for Communities & Local Government (DCLG), more or less in line with what the capital is thought to need. Nor does it mean that they are all necessarily in a position to be built out the very next day.

But, still, the number of plots approved for residential development in a given year has increased dramatically, by 48 per cent between 2011 and 2015.

Here’s the thing, though: this has not been matched by anything like a corresponding increase in building activity. Starts have risen over the same period by just 26 per cent, from 110,820 in 2011 to only 139,680 in 2015. If we’re ever to increase housing supply to the required levels, this growing shortfall between permissions and starts will need to be addressed and overcome.

Now, the Home Builders Federation (HBF) dismisses the starts figures as unreliable. And it is true that they probably underestimate activity somewhat  usually by a few thousand units, sometimes possibly by up to 20,000 units.

But even 20,000 (and I’m being really generous there) is nothing compared with the gap that has emerged between new planning permissions and building starts – a gap which was 65,389 in 2011 and, by 2015, have grown to a massive 121,964.

Sure, there is a little bit of give in the figures, but the trend is clear: homes simply aren’t being built as quickly as they are being approved by planners.

Another way of measuring this is to look at the completions numbers from the government’s other housebuilding data series, the more reliable “Net supply of housing”; this is published each November and is the HBF’s preferred source of figures.

Of course, completions take rather longer than start’s, so we have to allow for a degree of timelag: builders reckon homes take two to threee years to proceed from planning permission to being ready for occupation.

But even allowing for that, the same discrepancy occurs. There were 195,300 homes approved in 2012, for example; three years later there were still only 155,080 completions. Whichever we cut this, whether we look at starts or completions, it is clear that building activity has not responded proportionately to the rapid increase in land ripe for development.

The gap between permissions and building activity that has emerged since 2011 can be appreciated at a glance from the following graph. Note too that, pre-crash, while there was a gap back then, it was much tighter.

So what’s behind this? To be fair to developers, many of these planning permissions may not be in their possession. There has long been an issue of non-building landowners (speculators, historic landowners, public sector agencies, and so on) sitting tight while their holdings rise in value. This has been well documented in London by the consultants Molior.

However, Molior has charted a decline not an increase in the number of unbuilt planning permissions being held by non-builders over recent years, from 45 per cent in 2012 to 32 per cent in 2014.

So that still leaves two-thirds of unimplemented permissions in the capital in the hands of developers. So why are they not throwing up houses? Is there not a shortage of homes in this country?

Unfortunately the private housebuilding industry does not cater to housing need: it caters to effective demand. The building activity depicted in the graph above (minus 20,000 or so housing association starts each year) reflects not how many houses could be built, but how many willing buyers there are ready to purchase a new-build property at current market prices.

Developers do not build out sites as quickly as they physically can. They build them out as quickly as people are prepared to buy them – at the current market price or higher. As Philip Barnes, group land and planning director at Barratt Developments, recently observed:

The reality is that housebuilders, as return-on-capital businesses are not able to build our products at a pace faster than our customers will purchase them, at the market value.

“We could in theory cut prices to speed up sales – but we have based our land purchase price on the estimated market values so we don’t have this option in practice.”

Does this amount to landbanking? Developers say it doesn’t.

But, whatever we might call this process, the effect is that they are not releasing land with houses on it back into the market until it has reached the price they need to achieve to turn a profit. They are acting perfectly rationally in the present market – if I were a housebuilder I would do the same – but this is not an arrangement which is compatible with dramatically increasing the number of homes to plug a housing shortage.

Addressing this dynamic is something that Theresa May’s government must prioritise if it is to make any real headway on housing. There are various policies that have been uggested for tackling it. Opening up the market to SME builders by increasing the number of smaller development sites would be a good start, although it is unlikely this would make quite the difference that is needed.


A bolder route would be for the government or councils to directly commission builders (including, again, many SMEs) to put up the houses we need, and so bypass the private sector; but this will involve spending a good chunk of money.

Short of that, private developers will have to be incentivised to build more quickly. This will probably mean giving – and requiring – local authorities the power to impose contractual obligations about the pace of development when granting planning permission in the first place. It will mean creating a framework within which developers do not bid each other up for every plot of land to such a price that they cannot afford to build it out at speed.

If it is clear and explicit enough, the local authority’s requirements for all developments could be used by developers to signal to landowners that they cannot pay above a certain price for the land. Land prices, the repositary for most of the inflated value in housing, would be anchored.

Don’t get me wrong: we are still going to need lots more land, in perpetuity, to keep the housing pipeline going. This is not a Nimbys’ charter. It will still be a big task to ensure enough new homes are approved by planners – and in the right places – to meet the rapidly growing need for housing.

But the principal challenge right now is not increasing the numbers of new planning permissions. It’s getting those that are granted built out much more quickly.

Daniel Bentley is editorial director at the think tank Civitas and tweets @danielbentley. His briefing paper “Planning approvals vs housebuilding activity, 2006-2015” was published this week. 

Want more of this stuff? Follow CityMetric on Twitter or Facebook.

 
 
 
 

12 things we learned by reading every single National Rail timetable

Some departure boards, yesterday. Image: flickr.com/photos/joshtechfission/ CC-BY-SA

A couple of weeks ago, someone on Twitter asked CityMetric’s editor about the longest possible UK train journey where the stations are all in progressive alphabetical order. Various people made suggestions, but I was intrigued as to what that definitive answer was. Helpfully, National Rail provides a 3,717 page document containing every single timetable in the country, so I got reading!

(Well, actually I let my computer read the raw data in a file provided by ATOC, the Association of Train Operating Companies. Apparently this ‘requires a good level of computer skills’, so I guess I can put that on my CV now.)

Here’s what I learned:

1) The record for stops in progressive alphabetical order within a single journey is: 10

The winner is the weekday 7.42am Arriva Trains Wales service from Bridgend to Aberdare, which stops at the following stations in sequence:

  • Barry, Barry Docks, Cadoxton, Cardiff Central, Cardiff Queen Street, Cathays, Llandaf, Radyr, Taffs Well, Trefforest

The second longest sequence possible – 8 – overlaps with this. It’s the 22:46pm from Cardiff Central to Treherbert, although at present it’s only scheduled to run from 9-12 April, so you’d better book now to avoid the rush. 

  • Cardiff Central, Cardiff Queen Street, Cathays, Llandaf, Radyr, Taffs Well, Trefforest, Trehafod

Not quite sure what you’ll actually be able to do when you get to Trehafod at half eleven. Maybe the Welsh Mining Experience at Rhondda Heritage Park could arrange a special late night event to celebrate.

Just one of the things that you probably won't be able to see in Trehafod. Image: Wikimedia/FruitMonkey.

There are 15 possible runs of 7 stations. They include:

  • Berwick Upon Tweed, Dunbar, Edinburgh, Haymarket, Inverkeithing, Kirkcaldy, Leuchars
  • Bidston, Birkenhead North, Birkenhead Park, Conway Park, Hamilton Square, James Street, Moorfields
  • Bedford, Flitwick, Harlington, Leagrave, Luton, St Albans City, St Pancras International

There is a chance for a bit of CONTROVERSY with the last one, as you could argue that the final station is actually called London St Pancras. But St Pancras International the ATOC data calls it, so if you disagree you should ring them up and shout very loudly about it, I bet they love it when stuff like that happens.

Alphabetical train journeys not exciting enough for you?

2) The longest sequence of stations with alliterative names: 5

There are two ways to do this:

  • Ladywell, Lewisham, London Bridge, London Waterloo (East), London Charing Cross – a sequence which is the end/beginning of a couple of routes in South East London.
  • Mills Hill, Moston, Manchester Victoria, Manchester Oxford Road, Manchester Piccadilly – from the middle of the Leeds-Manchester Airport route.

There are 20 ways to get a sequence of 4, and 117 for a sequence of 3, but there are no train stations in the UK beginning with Z so shut up you at the back there.

3) The longest sequence of stations with names of increasing length: 7

Two of these:

  • York, Leeds, Batley, Dewsbury, Huddersfield, Manchester Victoria, Manchester Oxford Road
  • Lewes, Glynde, Berwick, Polegate, Eastbourne, Hampden Park, Pevensey & Westham

4) The greatest number of stations you can stop at without changing trains: 50

On a veeeeery slow service that calls at every stop between Crewe and Cardiff Central over the course of 6hr20. Faster, albeit less comprehensive, trains are available.

But if you’re looking for a really long journey, that’s got nothing on:

5) The longest journey you can take on a single National Rail service: 13 hours and 58 minutes.

A sleeper service that leaves Inverness at 7.17pm, and arrives at London Euston at 9.15am the next morning. Curiously, the ATOC data appears to claim that it stops at Wembley European Freight Operations Centre, though sadly the National Rail website makes no mention of this once in a lifetime opportunity.

6) The shortest journey you can take on a National Rail service without getting off en route: 2 minutes.

Starting at Wrexham Central, and taking you all the way to Wrexham General, this service is in place for a few days in the last week of March.

7) The shortest complete journey as the crow flies: 0 miles

Because the origin station is the same as the terminating station, i.e. the journey is on a loop.

8) The longest unbroken journey as the crow flies: 505 miles

Taking you all the way from Aberdeen to Penzance – although opportunities to make it have become rarer. The only direct service in the current timetable departs at 8.20am on Saturday 24 March. It stops at 46 stations and takes 13 hours 20 minutes. Thankfully, a trolley service is available.

9) The shortest station names on the network have just 3 letters

Ash, Ayr, Ely, Lee, Lye, Ore, Par, Rye, Wem, and Wye.

There’s also I.B.M., serving an industrial site formerly owned by the tech firm, but the ATOC data includes those full stops so it's not quite as short. Compute that, Deep Blue, you chess twat.

10) The longest station name has 33 letters excluding spaces

Okay, I cheated on this and Googled it – the ATOC data only has space for 26 characters. But for completeness’ sake: it’s Rhoose Cardiff International Airport, with 33 letters.

No, I’m not counting that other, more infamous Welsh one, because it’s listed in the database as Llanfairpwll, which is what it is actually called.

 

This sign is a lie. Image: Cyberinsekt.

11) The highest platform number on the National Rail network is 22

Well, the highest platform number at which anything is currently scheduled to stop at, at least.

12) if yoU gAze lOng into an abYss the abySs alSo gazEs into yOu

Image: author's own.

“For I have seen God face to face, and my life is preserved”, said Thomas.

Ed Jefferson works for the internet and tweets as @edjeff.

Want more of this stuff? Follow CityMetric on Twitter or Facebook.