The real problem behind the ongoing housing crisis is unresponsive supply, write Paul Cheshire and Christian Hilber. They explain that the solution is not hard to fathom: to build more houses, there first has to be land, and then an incentive to build on that land. This solution nevertheless requires radical change in the inertia of UK housing policy.
Since 2015, neither the government nor any opposition party has had any effective policy to tackle our housing crisis symbolised most starkly by Starter Homes. To the extent there has been action, it is largely ineffective at best and counterproductive at worst. All that has happened is that affordability has continued to worsen – and to worsen most where people most want to live.
The ability of the under 40s to buy a house has continued to be eroded and both inter-generational and inter-regional inequalities increased. The young have lost out because, as the Institute for Fiscal Studies has shown in a series of studies, they are systematically losing access or even, for an increasing proportion, any hope of access, to the single accessible asset that has done best over the last generation: and now increasingly even the ability to rent decent housing in places where there are jobs. Those in failing local economies in the North are locked into hopelessness because access to housing where there are good jobs is way beyond their reach.
Demand-side policies are popular but doomed to fail
Boris Johnson’s government and its immediate predecessors have introduced a number of housing policies, with Help to Buy the most hyped. This claimed to help younger lower income households get access to home ownership. Far most popular among the various Help to Buy offers has been the Equity Loan scheme. This provides a loan for up to 20% of the house value (or inside London, up to 40%) to buy new build properties. Buyers of such properties also do not pay interest for the first five years after purchase.
The intention may have been to help would-be buyers to overcome their credit constraints, but the effective outcome has been to push house prices up further, without any measurable impact on construction, in those markets that are least affordable and where the productive jobs are concentrated – such as Greater London. The trouble is that in these markets, housing supply is extremely unresponsive (we’ll get to that later), so when the government stimulates housing demand, this only makes existing houses even dearer, making existing home owners better off. And, by the way, there is one other group that benefits: the policy helped to push up the profits of developers registered in the Help to Buy business. While Help to Buy did increase construction in more remote areas where it is easier to build, these were not the places abundant in good jobs. The ultimate outcome is that young would-be buyers are even more ‘priced out’ in areas with productive jobs, and workers have to commute ever longer distances from remote areas leapfrogging Green Belts.
Other demand-side policies have very similar effects. Consider the Stamp Duty Land Tax exemption to first-time buyers: economists have long argued that the stamp duty damages both the housing and labour markets, and a recent study suggests that it has a particularly bad effect on housing-related and short-distance moves. It discourages downsizing and makes it more difficult for young families to expand their housing consumption. Phasing out stamp duty and reforming other property taxes would be the right policy. An exemption to first-time buyers mainly increases the price of smaller homes in the least affordable areas – think of London or Oxford – with the most unresponsive housing supply.
The real problem: unresponsive housing supply
Demand-side policies are politically easy but since the true driver of the housing affordability crisis is too little supply, which cannot respond to prices, such policies are inevitably ineffective. Supply problems are triggered by three causes:
- First, restrictions on land/space prevent cities from growing horizontally (Green Belts) or vertically (because of height restrictions and view corridors). Rebuilding at higher densities is also made difficult because of preservation policies, such as conservation areas or listed buildings.
- Second, local communities have virtually no incentive to permit residential development. There are strains on local services from the new residents but no revenue flow generated to provide infrastructure and public services.
- Third, the discretionary ‘development control’ planning system is dysfunctional, generating uncertainty (amplified by our system of Section 106 Agreements, supposedly to capture ‘planning gain’) and caters to NIMBY residents.
Playing with words
In measuring new supply, what matters is house building – completions – rather than the increasingly popular measure of ‘net additional dwellings’. This is popular with politicians because the greater the shortfall of new supply, the less the incentive to demolish obsolete houses. Instead, they are refurbished and sub-divided. Obsolete offices and industrial building are converted into sub-standard housing. So, the worse the shortage of houses is, the more the volume of net additional dwellings rises relative to actual building of new houses – completions. Looking at the data for completions reveals the scale of the problem starkly. In the 30 years 1959-1988, 7,449,160 houses were built in England: in the 30 years 1989-2018, only 3,328,850 were built. That suggests a shortfall of 3,120,310 homes – 104,000 a year – over the last 30 years.
We have not just been building too few houses: we have also been building the wrong sort of houses in the wrong places. Demand is for roomier houses in areas close to productive jobs. Satisfying that demand makes sense both for the welfare of people and for the productivity of the economy. Instead, we have been building relatively more houses where the local economy is slack (but there is lots of brownfield land) and population relatively slow growing.
Compare Barnsley and Doncaster with Oxford and Cambridge. If we look at population growth and house building, we find exactly the reverse pattern to that which logic would suggest: systematically fewer houses built in the more prosperous and faster growing local economies. In the nearly 40 years from 1980 to 2018, 56,340 houses were built in Barnsley and Doncaster combined but only 29,430 in Cambridge and Oxford combined. The pattern of relative population growth was almost the reverse: a 95,079 increase in Oxbridge against a 29,430 increase in the Barnsley and Doncaster pair.
The myth of ‘no shortage’
There is a narrative of ‘no shortage’ based on the fact that rents have been relatively stable over the past 15 years or so, even as house prices have risen. But rents, of course, are the price of housing services: not the price of houses as an asset.
The most important determinant of the demand for housing services is real incomes and these have barely regained the level they were at before the financial crash. Since the crash there has been an historically unprecedented fall in real interest rates, feeding through to low yields on all assets. That the fall in yields includes housing should surprise no one.
So rents not rising is not evidence that houses are in adequate supply. Rents are still unaffordable by historic standards, but the price of houses has ultimately been driven by falling yields, future expected real income growth and, crucially, their long-run shortage.
What should the next government do? A concrete policy proposal
The ‘no-shortage’ narrative is comforting, but it is fake news. And the only practical way to resolve our crisis is the uncomfortable process of building more houses. But in fact the process itself is easy: it is the politics that is difficult, especially when the payoff from success is well beyond the average politician’s time horizon.
A report for the Centre for Cities in 2019 showing how two million new homes could be built close to train stations serving major employment centres at no cost to government at all and at considerable gain to local communities where they were built. At present, incentives are so badly aligned that everyone resists new development. The community see eye watering capital gains made by a few land owners, but only costs to themselves because new development provides essentially no revenues but more residents to serve.
But if the owners of the stations (such as Transport for London or Network Rail) were given the sole right to develop, they could buy land cheaply – for not much more than agricultural prices instead of the £3.5 million per hectare minimum that currently has to be paid for housing land anywhere in southern England. If developing the land around stations were given incentives by tapering off subsidies to commuter rail and co-ordinated by a newly-created Development Corporation (along the lines of New Town Development Corporations created to address the housing crisis of the 1950s), it would ensure that the land was built out according to rail capacity and urgently. At the same time, the Development Corporation could ensure the new homes were built out in rail and cycle friendly fashion so current car-oriented developments were replaced by low carbon rail-oriented ones.
In addition, if current and extraordinarily inefficient negotiated planning obligations (Section 106 Agreements) and the Community Infrastructure Levy were abolished and replaced by a straight and completely predictable and transparent tax on the sale price of all the new construction, there would be proper funding to pay for enhanced services and infrastructure for the local communities receiving the development – place-making infrastructure – and proper funding for truly affordable housing in the social sector. Government would not have to pay a penny. All this could be funded from land value uplift.
The only catch is a political one. The policy would allow building on Green Belt land – so long as that land had no environmental or amenity value – within 800 metres of commuter stations. There is a remarkable quantity of such land. A worked example for five city-regions – Birmingham, Bristol, London, Manchester and Newcastle – reveals 47,000 hectares of buildable land. Reserving 10% for new publicly accessible green space and wild life habitat still provides enough land to accommodate two million additional homes – 15 years of building at recent rates. If the new development levy were set at 20%, it would produce £100 billion of revenue over time to pay for the place-building infrastructure and new social housing. But people could find new, more affordable and low carbon homes closer to where the jobs are.