Redesigning Housing Policy

Introduction: How to tackle the housing crisis

There is nothing new about criticism of ‘urban sprawl’ or proposals for new settlements. These two observations were made in 1945 and 1946:

Nor is the issue of housing affordability new, as this observation from 1973 demonstrates:

But what would a successful housing outcome look like?

What is a household? Does it matter that more young people live with their parents?

A perfectly-functioning housing system is unattainable. Housing is complex – a long-lived and spatially fixed product whose price is driven in part by price expectations. But nevertheless it is important to tackle the obvious failures:

Lack of stock at the bottom of the market. There are now over 5,000 rough sleepers in England, compared with fewer than 2,000 in 2010. The charity Crisis puts the number rough sleeping, or in cars, much higher, at 24,000. In London an estimated 1.6 per cent of households are in temporary accommodation – many of these including children.

There is a simple narrative about the housing problem. It runs as follows: the house price/income (HP/I) ratio has risen sharply (from around 5× in 2002 to 7.8× in 2018 for median homes and incomes in England and Wales), making home-ownership unaffordable. Greater housing supply would reduce prices. So all that has to be done is to build more, especially in fast-growing areas and their surrounding Green Belt, and all will be well.

The review of housing supply that I led for Government in the early 2000s contributed to this narrative to an uncomfortable extent. Politicians prefer it to difficult debates about housing taxation – and it provides a strong justification for resisting any local reluctance to accept new housing. And it is true that some of the suggestions below would make sense only in the context of a much greater supply of new homes – for example making housing benefit levels more generous. The housing crisis will not be resolved without rising to the supply challenge.

That said, it will not be a sufficient answer. First, the fall in long-term real interest rates has been a major driver of the increase in all asset prices, including housing, and seems unlikely to be reversed to any great extent over the coming decade. Pushing up new supply is not going to offset this price rise enough to make it easy for would-be first-time buyers to save for the deposits now needed. Second, housing problems manifest themselves very differently across the United Kingdom. The focus is often on home ownership in the South-East, but in the North-East it is unaffordable rent that is the key issue.

In the discussion of housing below, much of it refers only to England, as some policy is devolved. (The analysis also draws heavily on recent work by Glen Bramley for the NHF, and various publications from the excellent UK Collaborative Centre for Housing Evidence.) ²

Homelessness and social rent

In 1981, the stock of social housing in England was just over 5.5 million units, and the population of England was just under 47 million. By 1991 the social housing stock had fallen to just over 4.5 million, while the population had risen by around 2 million. The loss of stock under the Thatcher government was due mainly to the ‘right to buy’ policy – which helped those who bought, but removed the dwelling from re-cycling into the social housing stock.

Since then, the social housing stock has declined slowly, to around 4 million, a level at which it has recently more or less stabilised, but population growth has speeded up, with about 7.8 million more people. Only 41 per cent of social housing households are in employment, and 28 per cent of tenants are over 65 – social housing has tended to be available only to the most needy.

The cost of housing, relative to low incomes, has changed little since the 1980s, and there is evidence of unmet need both in the estimate from Bramley (ibid.) of 330,000 UK households effectively homeless (this includes temporary accommodation, squatting and those in unsuitable buildings), and in the rising numbers of those receiving housing benefit.

It is extremely hard to see a solution other than increasing the social housing stock and widening the scope of eligibility. Over the past decade, the social stock has grown by a mere 120,000. This sounds expensive – another million social homes may be needed over ten years – at a cost of perhaps £20 billion per year. However, some units may be built by housing associations using cross-subsidy from market housing development, and some units could come from a more aggressive use of S106 agreements. In addition, the housing benefit bill should fall.

The private rented sector (PRS)

In 2018, 4.8 million households in England, almost 20 per cent of the total, lived in the private rented sector (in the early 2000s the share was just under 10 per cent). Under the proposal above, some of the poorer households in this tenure could be expected to flow into the newly-enlarged social housing stock.

The housing benefit bill has doubled since the early 2000s, and is now £22 bn annually. During the 2000s, expenditure on housing benefit rose rapidly, due to a rising number of claimants, especially in the private rented sector (PRS). Since 2010, the government has prevented the housing benefit bill from continuing to increase by limiting the rents that it is prepared to underwrite. In the social rent sector, the pressure of this rent control falls mainly on the housing providers – but in the PRS where there are around 1.2 million housing benefit claimants (and where over £9bn of housing benefit is paid out), it is unclear how the outcome has been divided between lower rent for landlords and more stress for less well-off tenants.

A generally accepted view is that if a household is paying more than around 33 per cent of its income in housing costs, it is in housing stress. A recent estimate (Affordable Housing Commission, 2019) suggests that two million households in the PRS are in housing stress, and in particular that 40 per cent of those with below-median incomes are paying over 40 per cent of their income in housing costs, leaving often inadequate residual income. Rents on average have risen little in real terms since the 2000s, but they are up sharply relative to income for the lowest two income deciles, and also in London and other housing ‘hotspots’.

Would-be first-time buyers (FTBs)

The ‘affordability problem’ is often shorthand for the inability of the young to access home ownership (HO). In many locations this is now related less to the increase in the HP/I ratio and more to inability to raise a deposit. This has resulted from the combination of the fall in real interest rates, which has pushed up all asset prices, and restrictions on lending at high loan-to-value ratios in the interests of financial stability. Government policy has sought to address this largely through the much-criticised Help-to-Buy (HtB) Equity Loan, whereby the government lends part of the deposit on new-build homes, in return for a proportionate share in the change in value of the property.

Positive outcomes of this policy, which since 2013 has supported over 200,000 households into home-ownership, are: it has boosted new build supply; and it has enabled couples to move early into 3-bedroom houses, avoiding the need for the subsequent cost of the move as they stepped up from 2 bedrooms. (In 2006–7, 50 per cent of new build homes had 3 bedrooms or more: by 2017–18 the figure was 72 per cent).

However the policy is also open to four criticisms: it has enabled an increase in the new-build premium (though a recent NAO report found little evidence of this); it cannot discriminate against young people who might otherwise have benefited from family support; the government share may make it harder for households wishing to move further up the housing ladder; and it exposes the government to financial risk in the event of a housing market downturn (at a time when the public finances are likely to be under pressure).

It is expected that HtB will be phased out by 2023, with regional price caps introduced in 2021. It is already restricted to FTBs. However, the problem of couples who could afford to pay a mortgage but struggle to save for a deposit, unaided, in less than five years will not have been solved, even should house prices by then have fallen back a little relative to earnings.

Apart from family, the other players who might step in to support deposits instead – government, mortgage lenders, or developers – are themselves already vulnerable to a market downturn. A more radical solution, broadly supported by the Resolution Foundation, would be to endow all young people with a capital sum.

The housing supply numbers game

The question ‘How many new homes should we build a year’ cannot be answered in isolation from consideration of what the goals of policy are, and where the homes are built. It is also affected by many factors, including economic trends and migration flows.

The traditional approach to this question is based on the official household projections, which are not forecasts, but rather projections from previous trends. Actual annual household formation has been running well below these projections for at least the past twenty years – not least because weak housing supply means nowhere for households to form. (Household size was expected to decline, but has been compelled to stabilise.) In the 2004 housing supply review, I sought to shift the debate from these projections towards a more economic measure – how many houses would be needed in a region to achieve a desired HP/I ratio. As discussed above, I now think this over-simplistic.

More recently, the Ministry of Housing, Communities and Local Government (MHCLG) has sought to combine the household projections with a measure of how far the workplace-based HP/I is above 4× as a standard method of producing a target number for future new supply by local authority area. This proposal has many flaws, and was undermined in September 2018 when the Office for National Statistics (ONS) produced updated projections for household formation in England of 159,000 per year, down from 210,000 in the previous projection.

Use of this standard method then produced some odd results: for example, a big downward revision for Cambridge, due partly to the treatment of migration. Since then MHCLG have asked local authorities to continue to use the previous set of household projections as a short-term fix. And in principle this does have the merit of introducing an element of responding to demand for housing (rather than talking about ‘need’).

The simple use of household projections is unsatisfactory for a number of reasons (Barker, 2015) and, while the standard method aims to prevent long pointless debate about precise numbers, it can only ever be a starting point in any local authority. Given the evidence of previous undersupply, a different approach, adopted by Glen Bramley (2018), is to estimate the number of households in housing need due to present financial stress, including the problem of saving for a deposit. On this basis, and aiming to ‘solve’ the housing problems over 10–15 years, Bramley estimates that England should be building 340,000 new dwellings each year (380,000 UK-wide), rather above the present government target of 300,000 in England. (The latest actual is 222,000 in 2017–18.)

Bramley's estimate allows for an increase in both vacancies and demolitions, anathema to some, but an important part of housing market adjustment. However, the estimate might be thought high given that it is based on a higher set of population estimates than the latest ONS projections. Adjusting for this, a 300,000 figure seems a reasonable broad aim, if the premise is accepted that housing financial stress should be reduced significantly.

Some argue that environmental factors are a reason to constrain this demand. Certainly, issues of biodiversity and landscape need to be given full consideration in choosing sites across a local authority area, as does transport and the need to reduce energy use. Many of these issues would be addressed better by looking across wider spatial areas than just one local authority. Regional Spatial Strategies had a lot of merit. Each local authority can contribute to a regional plan, submitting its land allocations in a common format, rather than receiving a top-down allocation from MHCLG. And some are undeniable consequences – including the loss of some open landscape and the need to improve water supply if new development were to be heavily concentrated in South-East and Eastern England.

Infrastructure

Infrastructure discussion is often curiously confused. The demand for schools and medical care surely stems from the number of people and their ages in the UK, not the number of dwellings. In principle health and education plans should keep up: in practice this does not happen because schools, hospitals, and GP surgeries are ‘lumpy’ investments, and there is often a ‘crowded’ period before a significant new investment is triggered. This should be tackled by health and education spending policy.

Similar considerations apply to economic infrastructure (transport, energy, water, digital). The National Infrastructure Commission's (2018) Assessment looks in detail at the complex long-term issues here, although mostly it does not identify spatial conclusions. These recommendations should all be implemented, with government spending of at least 1.2 per cent of GDP on economic infrastructure achieved over the long term. Even then it is likely, in a relatively densely-populated island, that many areas will still press for additional funding to ease congestion.

Tax policy towards land used for residential development and the contributions to infrastructure and affordable housing are a topic of current debate. The answer seems simple in theory. When house prices (and therefore land prices) are high, it is necessary to find a way of taxing land that gains residential planning permission so as to maximise the tax take without unduly disincentivising landowners from bringing sites forward. This should also provide good financial incentives to local authorities for granting planning permissions.

At present there are a number of proposals aiming to achieve this by compulsory purchase (Aubrey, 2015), with compensation excluding ‘prospective planning permission’. That might be too draconian, and any change to the taxation system needs to allow for a transition, as a considerable quantity of land is always moving through development at any one time. Of course, were house prices to decline in real terms (as many wish), there would be less value to share around. Scope for this tax or levy would then automatically reduce; local authorities may share in the downside as well as the upside.

Place, inequality and mobility

A body of evidence suggests that the UK (and England) is unusually spatially unequal (Martin et al., 2016). (Not to forget that the country's major cities and more prosperous regions contain areas of extreme poverty.) It is striking that, notwithstanding a series of policies aimed at closing these spatial income gaps, they have tended to widen since the early 1980s.

Recent research from the US (Bayoumi and Barkema, 2019) and the UK (Judge, 2019) has outlined how the widening differentials in house prices between places have reduced labour mobility. This is not a new phenomenon, but has worsened recently due to the relative fall in earnings for the young, combined with the widening place-based income gaps. It is plausible that this lower mobility is affecting skills acquisition and productivity.

However, this is not a problem that can be resolved in the housing market, unless the only route is much higher home building in the presently successful areas (which is not being achieved; see Hliber and Vermeulen, 2014), and managed abandonment in the failing places. Rather, it is necessary to have stronger and more sustained policies aimed at strengthening economies across the country, enabling those living there to reach their full potential, and narrowing house price gaps to foster movement of younger workers in particular.

There have been many attempts at regional policies, but they have suffered from a variety of weaknesses: simple lack of resources; being spread too thinly; and being halted without having been given a proper chance for success. Place is one of the five foundations in the government's Industrial Strategy – but it will be at least a decade before it will be possible to judge the success of the policy initiatives. During this time it will still be important to supply sufficient homes in the successful high demand areas.

Conclusion

While the housing crisis stems in part from familiar issues around restrictive planning policy, it is also a symptom of problems in the wider economy. It will not be solved simply by fiddling with housing policy. These problems are: low relative wages for the young; poor planning of infrastructure; a lack of clarity about environmental issues; and failing places. These problems should be addressed directly. Further, the setting of monetary and macroprudential policy will not always serve the cause of housing policy at the same time as their inflation and financial stability targets.

In the housing market itself the issues are: the number of people who simply cannot afford decent housing; how to manage financial risk when house prices are high relative to incomes and an apparent inability to push up the supply of market housing sufficiently. The last reflects both problems in the acceptability of high housing numbers in local plans, and also the ‘scarring effect’ on the housebuilding industry of market downturns – each one of which has sharply reduced the numbers of small- and medium-sized housebuilders.

The housing problem also could be characterised as a problem of the distribution of space. Census data reveal that, since the 1980s, housing space has become more unequally distributed, at least in terms of rooms (Tunstall, 2015). However those who can afford space will purchase it – and the elderly often do not wish to downsize for a range of understandable reasons. It is unlikely that draconian policies to prevent this would be desirable, so under-occupation (relative to the ‘bedroom standard’) among owner-occupiers is likely to persist.

There are demand side fiscal issues too. The tax benefits of owner-occupation undoubtedly incentivise overconsumption of housing: a couple may stay in a large house when the family have left not just because they like the space, but also to continue to build up housing equity. The wealth gap between renters and home owners, who increasingly own second homes for their own use or as rental properties, has widened (Bangham, 2019), as has that between those owning homes in different parts of the country. Housing investment is viewed as too good a bet.

The outcome of all these trends is a knot of increasing inter-generational and inter-regional inequalities that are starting to damage social cohesion.

If it is necessary to spend more money to enable those who struggle to afford decent housing (and this paper is replete with suggestions to do just that), then it is necessary to find a way to raise more taxation from those who benefit from restrictions on housing supply, whether through reform of council tax, a wider wealth tax, or by introducing a limited form of Capital Gains Tax on principal residences. Without such an approach, some of the earlier policy proposals could prove counter-productive.

These tax changes are politically highly unpalatable. Yet the UK needs to start on this path urgently, not least because all the proposed policies engender significant transition problems. Once upon a time the abolition of mortgage interest tax relief was unthinkable. This new, supposedly unthinkable, taxation needs to be peddled over and over until it becomes a cliché, and politicians are able to act.