Mortgage Lenders, Barclay Card Centre Services, and the Crown Estate. ... Social polarisation in terms of council estates which are predominantly low income ...
Social Housing Forum
REINVENTING SOCIAL HOUSING FINANCE John Hills
Acknowledgments
The forum would like to thank the Joseph Rowntree Foundation for funding the writing and publication of this report. We would also like to thank our other funders for their support of the work o f the forum. These include the Housing Corporation, National Housing Federation, Pinnacle Housing Ltd, Council o f Mortgage Lenders, Barclay Card Centre Services, and the Crown Estate.
Contelits Preface Summary 1. Introduction
2. What are we trying to achieve?
3. What are the problems with where we are now? 4. Steps in the right direction
5. A more radical objective 6. A synthesis: back to the objectives
7. Conclusions
References
Preface IPPR established an independent Forum on the Future of Social Housing in July 1999. The forum has been set up to promote an open and inclusive national debate about the future of social housing, i t s role, its financing, its structure and i t s ownership. By focusing on a process of consultation and research, the forum aims to reach a consensual way forward for the social housing sector which i s visionary yet practical. Reinventing Social Housing Finance i s the first of a series of papers commissioned by the forum to encourage discussion and debate. The views
expressed are not necessarily shared by either IPPR or members of the Forum.
Sue Regan l PPR 2000
REINVENTING SOCIAL HOUSING FINANCE John Hills
Summary This paper argues that in designing social housing finance and housing benefit systems, potential reforms should be judged in terms of the fundamental aim of the system - to support those who would not otherwise be able to afford an acceptable standard of housing. This aim needs to be taken in the context of creating and sustaining communities and areas which include a social mix, and avoiding area polarisation. Some of the other justifications and explanations for the systems of subsidy which have emerged in the UK are weaker than they were in the past. A different range of instruments for supportingpeop1.e'~housing from those on which most emphasis has been placed in the past may now be appropriate.
Problems with where we are now Looking at the current systems, the paper identifies a series of problems: Social polarisation in terms of council estates which are predominantly low income and have high levels of worklessness. Polarisation in terms of the way in which people can make choices over their housing and have control over where they live depending on their tenure: these are far more restricted within the social rented sector than for the majority of the population. The quality and reputation of parts of social housing, in part related to the repairs backlog and to financial systems which have not supported adequate long-term maintenance of the council housing stock. This is a problem for those who live in it, but also means government has not been properly looking after one of its largest assets. A variety of value for money problems, both in subsidies for new housing in some areas where housing demand is low and in the effects of the '100 per cent subsidy at the margin' structure of Housing Benefit. Issues around the scale of operations which is efficient for day-to-day management and tenant responsiveness may be a great deal smaller than the scale which is efficient for capital finance purposes (but even that does not have to be huge). A bewildering series of conflicting principles underlying the levels of social rents varying within the council and housing association sectors as well as between them. This is not only an equity problem for tenants who do not receive Housing Benefit and pay different amounts for the same thing, but it also means that there is no way for tenants or others to use the level of rents as a consistent measure of the efficiency of their landlord. This anarchy is also a barrier to some reforms of Housing Benefit. Polarisation of the choices which people can make over the different components of the housing 'package' - such as there being no way to get into the housing market except by buying a whole property, or there being little support for low income owners with their repair and maintenance costs (in contrast to the help tenants receive for the equivalent costs via Housing Benefit). The attenuated state of the private rented sector, meaning that it is less effective in meeting particular life cycle needs than in other countries, and so pushes more people into marginal owner-occupation and increases the pressure on the social sector. However, the
high cost of the sector in its current state raises value for money problems where public finance is paying for rents through Housing Benefit. A variety of problems with the structure of Housing Benefit, including the feature that for all recipients, benefit moves pound for pound with any variation in rent (until a sharp cutoff is reached through ceilings on eligible rents). Given the rationing across much of the current social housing allocation system, this may not be too great a problem now. However, as part of a housing system which did have more choices and options for tenants or was intended to increase efficiency pressures on landlords, it would be bizarre. Potential solutions One solution to some of these problems would be to move towards a system which eliminated specific housing subsidies, left existing social landlords as non-profit bodies competing without subsidy in the market, and replaced Housing Benefit with flat rate additions to benefit levels and to tax credits and tax allowances. Starting from where we are, there would obviously be a very difficult transitional problem, with substantial individual gainers and losers from such a switch (even if the distributional consequences were neutral overall). These would be difficult to cope with, but more fundamentally this solution would not meet the key objectives set out above. The potential consequences of drastic area polarisation, overcrowding and substantial hardship for some are not attractive. This suggests that the 'pure market plus cash redistribution' model should be rejected even as a long-term aim. But this does not mean that we should stay where we are - with much housing support run on a 'food parcel7 approach, where people get little choice over their housing and few trade-offs between the cost to them and the quality or location of where they live. This approach already looks outmoded - with increasing affluence, it will look more so by the end of any realistic period for phasing in reform. Irrespective of decisions on the major issues of the overall level of social rents and the fundamental structure of Housing Benefit, Section 4 of the paper outlines some changes which would help reduce some of these problems. These include: Financial structures which help underpin resident empowerment. These include: organising and budgeting for day-to-day services locally, on a scale which may be much smaller than appropriate for capital operations; setting clear and consistent guidelines across the social rented sector for what day-to-day spending is expected to be and for how spending relates to rents so that tenants and others can judge efficiency (preferably between several landlords operating locally); and ultimately ensuring that tenants do have the power to 'exit' on level financial terms to another social landlord or to selfmanagement. Landlord financial structures which are clearly separated from other local authority finance (in the case of council landlords) and allow adequate resources to plan and provide for the long-term maintenance of the stock. In effect this means moving local authority housing onto a financial basis which looks much more like the current structure for housing associations. The paper suggests that this could be done through a series of stages, rather than only applying to those councils which opt for immediate outright transfer of their stock. The DETR's proposals for 'resource accounting' and for councils to retain a new Major Repairs Allowance are already a move in this direction. A corollary of this is the provision of enough resources from central government for council landlords to eat into the backlog of repairs. By way of return, central government would be reducing its ultimate long-term liability which this backlog represents. Like
local authorities, central government itself needs to move from financing social housing on a cash book annual basis to thinking of it in balance sheet terms. A set of consistent principles underlying rents and subsidy throughout the social sector. The paper suggests that one way of doing this would be through expecting rents to cover management and maintenance and provision for major repairs, plus a 'capital element' calculated as a percentage of current capital value. The size of this capital element which determines the degree of subsidy in economic terms - is discussed further below. A start towards the introduction of 'shopping incentives' within Housing Benefit. At the most modest end, this could involve fixing benefit for a period, rather than changing it every time circumstances change. If there were moves towards a 'hybrid' structure of benefit involving a flat rate element and only a proportion of actual rent, the paper suggests that it would be better to have a national flat rate element, with the proportion of rent used varying locally depending on average local rents in the relevant tenure (rather than to use a national percentage of rent with a locally varying flat rate element). An expanded strategy for a range of housing choices which are less sharply differentiated between tenures and locations. These could include: continuation of new mixed tenure developments with some low cost home-ownership; advertising and renting out properties in some currently solidly council rented areas at market rents; equity sharing offers for existing properties rather than just in new developments; use of resources from these to diversify the location of social housing; and opportunities for owners to sell part of the equity in their property to meet various needs, for instance in old age. A wider spread of equity sharing and the proposed structure for social rent setting would open up the opportunity for a balancing market in 'housing bonds', where people could buy a financial asset whose value was related to future property prices, without having to become a full owner of a particular property. A more radical objective These ideas would help create more mixed communities, greater tenant voice and influence, long-term viability of social landlords, coherence in social rent structures, and less polarisation in housing choices. They would have significant advantages in blunting some of the problems with existing arrangements, but they could not be claimed to eliminate them. Section 5 of the paper therefore examines whether it would be possible to set a much more radical long-term objective - moving social housing as a whole to a regime of low or zero subsidy in economic terms to reduce rationing and improve choice and competition - while protecting existing tenants. At the same time Housing Benefit could be moved towards a system with a significant national flat rate element. Care has to be taken with this, if it is not to set up incentives for new households to form or for phantom tenancies to be created to take advantage of it. To avoid this (but at some cost) it could be paid as an addition to all benefit rates, not just to tenants or to householders. As far as low income owners are concerned, a flat rate element eventually building up to £15-20 per week for a typical household could be justified on equity grounds as giving equivalent help with repair and maintenance costs to that already received by low income tenants. To protect existing social tenants, the paper suggests a system under which they could be granted an equity share in the property in which they live, reflecting the value of their current expectation of continuing to pay sub-economic rents. Landlords would then 'pay' them a return on these equity shares which would discount rents for this group back to current levels. This would make explicit what are currently implicit rights, while allowing the existing
structure of subsidies to be 'grandfathered out', and opening up a transformation in mobility possibilities by allowing trading in equity shares. This suggestion solves some of the problems associated with moves towards phasing out general subsidies to escape fiom rationing and allow much more choice and flexibility within what is currently social housing. However, it does not solve the problems for potential new tenants with low incomes, particularly those living in high cost areas. If Housing Benefit took the strain, the result would be much deeper unemployment and poverty traps. But a system of granting equivalent equity shares to some or all new entrants would leave us back much where we started, with a continued need for heavy reliance on rationing.
A synthesis: back to the objectives The question becomes whether there is an intermediate route between just modest rationalisation of the current system and such a radical transformation. Section 6 of the paper suggests that the key issue remains affordability of 'decent' housing by those with low incomes. We do not want to leave things entirely to the market because of the dangers of very low quality housing and extreme area polarisation which would result. We do not want to load everything onto the Housing Benefit system because in some areas this would cause very deep unemployment and poverty traps. Flat rate allowances as a replacement for a large part of Housing Benefit in high cost areas would create hardship for some. But these problems have one thing in common: they all relate to the difficulties in high cost areas and regions. They do not rule out a system which has lower levels of economic subsidy and greater relative reliance on a flat rate element within housing benefits in low and medium cost areas. As far as two of the key decisions are concerned, this suggests the following long run objectives: Within the proposed system for setting rents - where guideline rents for council housing and capital subsidy for housing associations are based on achieving a set net return on current capital values (after other costs) - the required return could vary depending on the value of the property. This could involve a 'low subsidy' required return up to the national average capital value of social housing, but a return implying a higher level of subsidy for the excess of values above that. Incorporating a nationaljlat rate housing costs element into benefit rates and tax credit structures, with locally varying percentages of rent covered by explicit Housing Benefit. In both cases it would be sensible to approach these changes in stages - say over ten years not just because this allows adjustment on the part of tenants, but because it also allows experience with the effects of the changes to build up, and if there are problems to adjust what is being done. The end result of both moves would be a logical solution to the fundamental objectives of policy and the constraints under which it operates. General subsidy to social tenants (in economic terms) would be modest in low and medium cost areas, allowing much more fluidity in the housing system and more realistic opportunities for choice and mobility. But subsidy would remain high in areas (including higher cost parts of low cost regions) where it was needed to allow a social mix and to avoid severe labour market problems. Housing Benefit would no longer be based on 100 per cent of rent in any region, so there would be 'shopping incentives' across the country - the corollary of more choice. However, even in the lowest cost regions low income households would still be receiving significant
specific assistance with their actual costs. In percentage terms, this specific assistance would be highest in areas where affordability problems were greatest. As a corollary, the need for the crude limits set by local reference rents, the single room rent, and the rent caps for councils and associations would be reduced, if not eliminated. Finally, however, it should be noted that these two moves have different distributional effects. The Housing Benefit reform could be introduced with no net distributional effects the average case in each area could be left unaffected But an increase in the capital element of target rents in the way suggested could mean a rise in council rents and real losses for existing tenants who were not receiving Housing Benefit. The paper suggests a way of protecting them against this using the 'equity share' ideas described above, but at a more modest scale.
Conclusion Running through the paper is a theme that we are still prisoners of decisions taken in the 1940s. These decisions may well have been appropriate solutions to the housing problems of fifty years ago. Indeed, in pointing to contemporary problems it is important not to forget the huge improvement in housing conditions to which they have contributed. But that very success means that the nature of the problems has changed, and with them appropriate solutions. In setting objectives we have to look at the needs of Britain in 2010, rather than just focussing on 2000. This need to look at reforms which are geared to a long-term goal has a political downside: the benefits will also arrive in the long term while some of the upheavals required will come more quickly. But if we never make the first moves towards a system better attuned to contemporary needs we shall forever be giving the answer that to get there we should not be starting fiom here.
The ways in which we pay for social housing2 and help those who live in it and elsewhere with their housing costs through Housing Benefit have a direct effect on more than a quarter of the population. As these households are drawn disproportionately from the poorer half of the population, any changes to existing systems have important distributional effects, not just in financial terms but also in terms of the choices and opportunities open to people. The way the financial systems operate also has major effects on the way a quarter of the national housing stock is run, and the condition in which it is maintained. The purpose of this paper is to examine key features of how these systems work at present and to suggest possible ways in which they might be reformed to underpin wider objectives. The paper does not try to be definitive: a range of ideas is put forward to provoke discussion, particularly within the contexts of the Institute for Public Policy Research's Forum on the Future of Social Housing and the debate around the Government's Green Paper on housing, promised for early 2000. One of the difficulties with the subject matter is that its ramifications stretch widely in several different directions. Decisions on Housing Benefit affect the structure of the rest of the social security system, the new tax credit arrangements for those with low incomes at work, and policies towards helping people make the transition into work more generally. The financial structures of social landlords are intimately involved with their management systems. Rent structures and the Housing Benefit system interact with allocations and letting policies. Capital subsidies for new developments and renovation of existing housing are important parts of more general regeneration and area improvement policies, and link closely to large parts of the current strategy towards countering social exclusion. Policies towards social housing are central to overall city policies, and to ideas of an 'urban renaissance'. It is hard to discuss finance for social housing without considering the treatment of the alternatives owner-occupation and private renting. In a comparatively short essay most of these connections cannot be pursued in detail.3 The core of the focus here is on subsidies to social tenants - the systems which mean that the rents they pay are below those which would be charged in the market,4 together with the Housing Benefit system, which pays all or part of the rents of low income tenants. It discusses the implications of these for landlords as well as for tenants, and hence for the 1
The author is Director of the ESRC Research Centre for Analysis of Social Exclusion, London School of Economics. The author is very gratehl for helpful comments on an earlier draft of parts of this paper from Stephen Duckworth, Paul Gregg, Chris Holmes, Alex Marsh, Anne Power, Peter Redman, Sue Regan, Christine Whitehead, Steve Wilcox, and members of the IPPR Forum on the Future of Social Housing. These led to many improvements, but the author remains responsible for any remaining errors and for all of the opinions expressed, which are not necessarily shared by those kind enough to comment. 'Social housing' is used here to include council housing and that owned by non-profit housing associations, now officially known as Registered Social Landlords, RSLs. For a more general examination of the place of housing subsidies within the tax and social security systems, see I-Iills (1991). For a discussion of the system incorporating more recent reforms, see Gibb and Munro (1999). For an examination of the ways in which subsidies to council housing and housing associations operate alongside the perceptions of social landlords themselves, see Glennerster, Hills and Travers (forthcoming). For more detailed discussion of Housing Benefit, see Kemp (1998,2000). For discussion of recent trends and a wealth of statistical material on housing finance in the LK, see Wilcox (1999). Note that 'subsidy' in this sense may exist even if there is no current financial flow from central government to a social landlord. For instance, a local authority may not be receiving any net contribution from Housing Revenue Account Subsidy, but still be charging sub-market rents thanks to low debts and the accumulated advantages of past financial flows.
effective use of the housing stock. It also looks at some alternative ways of assisting those with relatively low incomes with their housing costs. It does not discuss in any detail options for allocating or letting social housing, except to the extent that existing financial arrangements are bound up with current rationing systems. Nor does it discuss area regeneration policies or the wider structure of the social security system beyond the dilemmas associated with the Housing Benefit system. When looking at the position of social landlords in particular, the discussion mainly relates to the situation in England, rather than other parts of the U K . ~ The paper does not contain the detailed exemplifications, costings and distributional analysis which would be needed for a full appraisal of options for reform what it is intended to do is to look at the principles which underlie the finance of social housing and more general financial support for housing. Section 2 of the paper stands back from the current position to examine what the purpose of the whole apparatus is supposed to be. Section 3 looks at a series of problems with the systems which currently apply when set against such objectives. Section 4 puts forward some incremental changes which could help rationalise current systems and reduce the scale of some of these problems. Section 5 sets out some of the changes which might be both required and allowed by moving towards a much more radical change in the roles of social housing and Housing Benefit, while Section 6 suggests a synthesis of the previous two sections designed to help meet the fundamental objectives set out in Section 2. The final Section presents some conclusions for further discussion.
'
Subsidies to council housing in Scotland differ significantly Com those in England, for instance through rent rebate subsidy not having been amalgamated with general subsidies as it was in England. In addition, the moves towards 'resource accounting' for council housing discussed below are currently happening in England, but not elsewhere in the LTK.
2. What are we trying to achieve? While a wide variety of aims has been put forward to justify or explain state intervention in housing in the form of subsidies to social landlords or Housing Benefit for tenants with low incomes, the core aim which remains most relevant is to support those who would not otherwise be able to afford an acceptable standard of housing. In its traditional formulation, this is the aim of 'a decent home for all at a price within their means'. If this is accepted, there are several points which follow: The aim involves the combination of two factors: there being a minimum standard set by society; and there being a distributional aim - the core target for support is those with low incomes, rather than housing consumption in general. The desire to achieve minimum standards is the main reason why we don't simply redistribute cash via the tax and social security systems and leave people to buy their own housing in the market: faced with a market choice, some people with constrained resources would opt for a very low standard of housing (over-crowded, low quality) and higher consumption of other kinds. Implicitly this involves some form of paternalism: to protect other household members (e.g. children), or neighboursl the neighbourhood, or people's own future interests, we are prepared to support a different consumption pattern with higher housing costs than people would have chosen themselves given the cash. A related argument is that housing is a 'merit good' - voters are more prepared to support redistribution to the poor if it is in kind (health services, education, housing) than if it is in cash. While not necessarily inherently desirable, specific housing support may offer a more acceptable route to achieve a more equal society than cash transfers. If these last two arguments are unpalatably paternalist, the implication is the long run aim should be to move all redistribution onto a cash basis (although there might still have to be a prolonged transition period away from specific subsidies to get there while avoiding hardship given our starting point). In the past, the drive for minimum standards has taken different forms - getting rid of overcrowding, ensuring basic amenities, getting rid of slums, ensuring safe environments, etc. Those minimum standards have risen over time - but not as fast as overall living and housing standards. The proportion of the population who can now afford an acceptable standard of housing without specific support is higher than in the past. With rising affluence one would expect this proportion to rise further. Looking 15 years ahead one might hope, for instance, for an increase in living standards of the poorest by a third - or in less time if the Government succeeds in reducing relative poverty. This would, for instance, take someone who is currently in the middle of the second tenth of the income distribution to a real income equivalent to the current income of people near the top of the third tenth of the distrib~tion.~ In other words they (and people with similar incomes) would move from a position where they are very likely to receive housing support to a real income where today they would be most likely to be expected to get by without specific housing subsidies (although, of course, rising minimum standards andlor real housing costs slow this process).
Incomes (adjusted to be equivalent to those of a childless couple) after housing costs were about £1 12 per week in at the 1 5 ' ~percentile of the distribution in 1997-98 (DSS, 1999, figure 2.1, AHC). An increase of one third would take them to about £1 50, close to the current 291hpercentile.
If affordability of acceptable housing is the core reason for support, that support may only be required at certain points in the life cycle - for instance, parents with younger children and only one earner, or in the aftermath of divorce or partnership breakdown, for young people without the option of staying with their parents, or for low income pensioners rather than the assumption being that support is needed in perpetuity (for instance through being allocated a permanent subsidised tenancy with succession rights and the Right to Buy at a discount). This aim does not in itself imply any particular form of support out of the possibilities subsidised direct provision, housing benefits, subsidies for house purchase, etc - although some may give better value for money than others. There are vulnerable groups for whom direct provision may be justified on the grounds that they could not otherwise cope with the (non-trivial) organisational burden involved in being an owner or finding a private tenancy. However, this group is much smaller than those currently living in social housing. As one measure, the 1998 British Social Attitudes survey reports just 10 per cent as saying that being an owner-occupier would be 'just too much re~~onsibility'.~ Nor does this aim imply the need for general support for housing consumption, although this has been a both a deliberate and accidental outcome of policy in the past (for instance, when owner-occupation was much more heavily favoured through the tax system or when housing in general was favoured through the lack of any general tax on housing consumption in the era of the Poll Tax). However, it does not necessarily imply that support should only be for those with low incomes - if, for instance, support for others would help create mixed communities (see below) or could be justified in labour market terms (in reducing the unemployment or poverty traps) and there were no better alternatives. ving said all of which, there are three riders to this as the core aim of housing subsidies:
An increasingly important objective is the need to combat social exclusion by creating and sustaining communities and areas which include a social mix. It is not acceptable to meet the aim of affordability through supporting housing which is only in low income ghettos, or only in certain parts of the country. To put it the other way round, we want to ensure that some people with low or modest incomes can afford to live in London, the South East, outside the poorest parts of our towns and cities, and in rural areas. This means that support with housing costs will be needed for those with higher incomes in high cost areas than elsewhere. It may also mean trying to avoid incentives for people to leave particular areas as soon as their circumstances improve. The objective has to be met within what are inevitably limited public resources. This implies that the systems used have to be efficient in the sense of providing value for money to both individuals and the public purse. As discussed below, the current mix is not necessarily successful in doing this, but some of the alternatives could be even worse. Access to housing is not guaranteed just by financial mechanisms by themselves, but access routes for the vulnerable are required if the aim of a 'decent home for all' is to be met.
Figure from secondary analysis of the 1998 British Social Attitudes survey. As Ford and Burrows (1999) show, attitudes to home ownership - while still overwhelmingly favourable - became rather less enthusiastic after the house price collapse of the early 1990s. However, the proportion agreeing strongly that 'owning your own home is just too much responsibility' was only 3 per cent in 1998, an insignificant change from 4 per cent in 1986.
Weaker arguments With these exceptions, some of the other arguments which have been used to justifl housing subsidies and their particular form in the past look much weaker than they once might have done. It is useful to run through how things have changed, because several of these arguably obsolete arguments - explain how we reached our current position. In the past, particularly after the Second World War, a key reason for direct provision of subsidised (council) housing rather than using other routes was to ensure that available resources went into increased supply, rather than simply bidding up rents or land prices. This argument is far less strong given the better overall balance between housing supply and households since the 1970s (although it is still an issue in some high demand areas). At the same time, in the past the private sector offered much less of an alternative. A far smaller proportion of the population was in a position to afford the investment component of owner-occupation than now. The bottom end of the British private rented sector has, through most of the twentieth century, offered poor quality and low reputation accommodation. This is not, however, an inherent characteristic of private renting as a sector, as witnessed by its role in other countries. The gradual emergence of a niche sector of commercial private renting offers some hope of a healthier private rented sector in the future, absorbing some of the demands which would otherwise be thrown back on social housing or marginal owner-occupation. An important justification for setting social rents well below market levels in the past was that, without this, social tenants - generally poorer - would be in a far worse position than owner-occupiers, whose long-run costs were substantially reduced through favourable tax treatment. This has changed for three reasons: lower inflation and hence borrowing costs which are much closer to real interest rates; the withdrawal of mortgage tax relief; and the much less unfavourable treatment of other assets which are widely available (like Individual Savings Accounts, ISAs). Technically, the real 'cost of capital' to owneroccupiers was negative in real terms in the late 1970s, was around 1 per cent in the 1980s, but is now pretty well equal to the gross market real (after inflation) interest rate (say, 3.5 to 4 per cent).8 In the past, low council and housing association rents were needed in order not to violate fairness or 'tenure neutrality'. This is not currently the case (although that does not mean that rents should not be subsidised for other reasons). With a much less developed mortgage lending market and a smaller proportion of the population in owner-occupation, it could have been argued in the past that there were capital market failures in terms of the ability of those with moderate incomes to borrow against their future incomes. This argument is hard to sustain today (indeed, the danger is probably the other way, with people able to borrow more than one might think prudent). Housing subsidies have been given a number of macroeconomic justifications in the past: for instance, housing investment as a means of reflation and reducing unemployment without sucking in imports; or housing subsidies as a way of keeping down overall inflation. Whether or not these were good arguments in the past, they do not chime well with contemporary economic policies. Arguments for general housing support on the grounds that there are 'externalities' benefits to neighbours - from keeping property in good repair and so on are much weaker given the level of overall housing standards, although they still have force in particular neighbourhoods (and underlie a large part of regeneration activity).
Hills (1998), p.178.
Similarly, it has been argued that we need to subsidise 'building for the future', to allow people to meet standards which would not necessarily be chosen today. As a general argument this seems rather weak given that many aspects of housing can be improved later on, and there are other competing ways in which we could equally argue for investment for future needs.9
Strengths of owner-occupation Finally, in thinking about reforms to housing finance, it would be foolish to ignore the strengths of the dominant sector - owner-occupation - as a form of housing provision. Its popularity in the UK may well have been helped by its favourable past tax treatment and by the 1980s political rhetoric emphasising ownership as an aspiration. It has also had the advantage for consumers of a relatively well-functioning purchase market while the private rented sector has been in a mess, and social housing has been rationed. But some of these advantages have been eroded1' and it remains popular - as witnessed by the 89 per cent in the 1999 poll commissioned by IPPR who would want their children or grandchildren to be owners in 20 years' time, or the 79 per cent who would advise a young couple to buy, or the 84 per cent of British Social Attitudes survey respondents who argue that buying is cheaper than renting in the long run. There are four key inherent advantages to owner-occupation compared to any kind of renting: occupier control of changes and improvements; absence of some of the management costs inherent in the landlord-tenant relationship and the ability to carry out other kinds of management oneself (although some prefer to pay someone else to do it, achieved by being a tenant); internalisation of the benefits of looking after a property well; and the fact that owning allows one to buy one's future housing at a known price now (in financial terms, 'hedging' against the risk that it might cost more in the future). Of course there are downsides (some of which are discussed below), but it is not only because of the peculiarities of past British housing policy and the problems of other sectors (such as the difficulties for social tenants in moving house) that owner-occupation is so popular, and this popularity is an important part of the context for thinking about long-term social housing finance reforms.
However, this argument does re-enforce the case against building new homes which do not even meet today's standards, and can apply to basic infrastructure. It also acts as a justification for some of the minimum standards imposed through regulation - for instance to encourage new building of 'lifetime homes' - all of which have some impact on affordability, and hence potentially on the need for subsidy of those with low incomes. 10 Effectively, owner-occupation is now tax-free rather than positively tax-subsidised, putting it in a similar position to ISAs, a slightly less favourable position than pensions, but still a better position than tax-penalised private renting. 11 Figure from secondary analysis of BSA. Ford and Burrows (1999) show that 47 per cent of the sample now agree strongly with this statement, up from 44 per cent in 1986 and 30 per cent in 1991.
3. What are the problems with where we are now? Social polarisation The housing system and its financial structure as it is now are associated with two forms of social polarisation: On the ground, in terms of estates which are predominantly low income and have high levels of worklessness. This is not just a problem for social exclusion, but is also one of failure to meet what people want and their aspirations for the future (e.g. only 6 per cent in the IPPR poll wanted to live in 'mostly rented' areas - let alone exclusively rented areas). This geographical polarisation results from the combination of tenure polarisation, the way in which much social housing is in large single-tenure estates, and the way allocations have been made within social housing. It would after all be possible to have social housing which was only occupied by those with low incomes, but was pepperpotted, without it causing geographical polarisation. It is ironic that some of the social provision designed to avoid the kind of polarisation which one would expect in a free market has in fact generated polarisation. The current system is also associated with polarisation in the way in which people can make choices over their housing and have control over where they live. Social tenants receive a rationed allocation of housing with a fixed quality and price. As a result of the Housing Benefit system, for many this is free and for most variations in price at the margin do not matter. Ability to move between areas is highly constrained (except through very occasional transfers, or through leaving and losing the subsidy from the social tenancy, or through exercising the Right to Buy and then selling later on). Tn terms of a basic aim of access to affordable housing, this usually achieves its aim for those within the system. The same could be said of supplying food parcels instead of cash to benefit recipients, but that would hardly be acceptable in terms of denying people the ability to make their own choices about what they eat. With the mass housing problems of the past greatly reduced, the 'food parcel' approach to subsidised housing also looks increasingly unacceptable, and will become more so with rising affluence. Of course, food is different - similar quality food can be bought at much the same price in most parts of the country. The same is not true of housing - but coping with this 'problem of rent', unsolved since Beveridge, should be one of the long-term aims of any far-reaching housing finance reform. The quality and reputation of social housing Part of the poor reputation of parts of social housing is to do with the polarisation problems described above. But part reflects its physical condition. Not enough money has been invested in the local authority housing stock to maintain its condition. As a result a substantial repairs backlog has built up. This is not just a matter of the trade-offs which have been made between low rents and low quality in some areas, or constraints on the volume of resources which have been allowed to local authorities; it is also to do with the way the finances of local authority housing have been run, effectively on an annual 'cash book' basis, rather than as relating to a long-term asset: Rather than being able to put aside money from rents each year to cover depreciation (or equivalently to carry out or provide for periodic major repairs), councils have been expected to come back to central government to get permission to borrow (credit approvals) every so often to catch up with their major repairs.
Once a council is allowed to borrow, subsidy is adjusted to cover all the costs of the new borrowing. Councils have every incentive to maximise the scale of their repair problems when arguing for credit approvals. At the same time they have had no way of planning for them on a long-term basis. This contrasts with the financial position which housing associations now find themselves in, where planning for major repairs and their financing is now crucial, and where trade-offs can be made over the timing of repairs, maintenance and initial building standards. The Department of the Environment, Transport and the Region's 'resource accounting' proposals12 have the potential to allow councils to move more in the same direction by allowing retention of a new Major Repairs Allowance (see below). All this matters for three reasons: The disrepair problem means that many council tenants are having to live in low quality housing and environments. The Government has not been taking proper care of one of the largest assets for which it is responsible. h effect, it is guilty of a failure of stewardship of its assets. Ultimately, neglect means loss of the asset altogether. This is not just a problem for government; it also means the loss of part of the existing housing stock, and the need for a new house somewhere else, increasing pressure on green-field sites and making it even harder to meet future housing needs. Unless the quality - in terms of what people value, which may have as much to do with environment as that of individual dwellings - is high enough, people with any kind of choice will not live in council housing. This increases the social exclusion problems resulting when the only people living in an estate or neighbourhood are there because they are on benefit and have no choice. The value for money of current use of public resources Value for money problems emerge in several parts of the system, but take different forms in different parts of the country. Giving valuable capital grants to housing associations to build houses in low demand parts of the country which then have to be knocked down is the most acute - if rather rare - manifestation. But questions also have to be asked about the value for money of significant capital grants for new developments which are then let at rents which 'are only attractive to those on Housing Benefit'. When those who are making choices with their own money prefer an unsubsidised private tenancy to the subsidised housing association one, something is going wrong. People are showing that they prefer something where they have some choice over its location, for instance, to what is on offer from the social sector despite its subsidy. This does not suggest the subsidy is being well-spent.
But this is not only a problem in the social sector. Again, the most acute problems of poor value for money may be seen in private landlords charging high rents to the Housing Benefit system for properties they have acquired for very little in low demand areas. But there is a more general problem in high demand areas. Private landlords currently receive what look like very high rates of return on their investment (partly reflecting the risks they are taking) far greater than the cost to government (or owner-occupiers) of borrowing. For long-term tenancies financed by Housing Benefit, paying private landlords rents which give real returns
l2
DETR (1998, 1999a,b).
in some regions of over 9 or 10 per cent13 (over 6 or 7 per cent after allowing for maintenance, repair and depreciation) before allowing for capital gains does not look a very good bargain for the public purse. Bed and breakfast and some private leasing deals look even worse value. Until the last two to three years the '100 per cent subsidy at the margin' structure of Housing Benefit has also reduced efficiency pressures on social landlords. The combination of uncapped Housing Benefit, freedom to set assured tenancy rents, and reduction of capital subsidy for housing associations in the first half of the 1990s led to spiralling association rents. In effect, associations faced no 'hard budget constraint' of the kind businesses usually face through the market (or which public organisations face through fixed budget allocations). Both associations and councils do now face hard budget constraints through varieties of 'rent capping'. This restores efficiency pressures on them - but at the cost of even greater anarchy in rent structures (see below). Other areas where one might question the value for money of current arrangements include: The scale of organisation of some social landlords. There is little evidence of 'economies of scale' in housing management beyond a few thousand units - perhaps the reverse. On the other hand capital financing is cheaper for larger organisations (although even this does not imply that they need to be huge with more than, say, 10,000 units). It may be more efficient to run day-to-day housing management and repair on a much smaller scale than landlord capital finances, quite apart from the advantages of smaller scale in terms of tenant control and responsiveness. The disadvantages of combining local authority strategic and 'enabling' roles with the landlord role. In recent research looking at finances for social housing, primary health care and school^,'^ all the measures which gave clearer budgetary devolution to providers appeared to be welcomed as helping efficiency - including moves like 'ring fencing' council Housing Revenue Accounts which had been opposed by parts of the housing world at the time of introduction. In the case of council housing the confusion of roles has been muddied even more in their finances by including rent rebates in the calculation of subsidy - although this is planned to go under current DETR proposals. Value for money in capital repairs when councils do receive credit approvals, given the lack of net consequences for their finances. One could also query whether associations know whether they are receiving value for money for their major repairs spending, given the lack of benchmarks of the kind which exist for recurrent spending. The high generosity of some Right to Buy discounts - although, as discussed in Section 5 below, discounts should be measured against the capitalised value to a sitting tenant of paying sub-market rents in the future, which can be considerable. The structure of social rents Social rents are now subject to a bewildering array of conflicting principles through which their levels are determined, most of which bear little relationship to the value of the housing being provided. Discussing the current system with a range of social landlords in 1998 not one could identify the equity aims of the current financing systems - indeed most were l 3 The Joseph Rowntree Foundation index of private rents shows gross yields in Great Britain averaging 8.6 per cent in 1997 (Wilcox, 1998, table 52). l 4 Glennerster, Hills and Travers (forthcoming).
baffled at the idea that fairness or equity had anything to do with housing subsidies at all!15 This is in stark contrast to other sectors like the NHS or education. It is, however, hardly surprising. Abstracting from the effects on rents of varying efficiency or from rents agreed by housing associations in negotiating funding for particular schemes, the rents paid by social tenants in England can reflect at least seven different underlying financial principles: Underlying subsidy to council housing are 'guideline rents', based on a combination of relative capital values and regional earnings levels. This is what is supposed to matter for council tenants. However, movements towards basing subsidy on these guidelines have been subject to complicated 'damping arrangements' throughout the 1990s - what matters most for some tenants is what their council's rents happened to have been in the late 1980s. More recently 'rent caps' - designed to limit rent increases largely paid for by Housing Benefit' - depend on what rents happened to have been in March 1996. A council which had high rents at that time has more freedom of manoeuvre than one which had kept its rents down at that point. Housing association tenants in place by January 1989 have their 'fair rents' set by rent officers under somewhat obscure principles which are subject to pressure both from central government and the landlords themselves. Later association tenants have their 'assured tenancy' rents set by the associations themselves, reflecting the history of capital grants they have received (which can be traced back to a Housing Corporation model of rent levels which would be 'affordable' in different parts of the country), the amount of development they have chosen to undertake, and the association's previous financial strength. These rents are generally higher than fair or council rents. In parts of the country they are close to private rents. However, associations have also been subject to rent-capping in the last few years, with maximum increases linked - if associations want to continue developing - to the RP1 from where their rents were in 1997. Again, this freezes rent relativities where they happened to be when the music stopped rather than relating to the value of accommodation being provided. Finally, 'transfer associations' generally give their existing tenants a guarantee on future rents not increasing more than certain amount more than inflation from where they were at the point of transfer. New tenants have their rents determined by the transfer terms (price or dowry), reflecting those assumed in the DETR's transfer valuation model (which may or may not be consistent with any of the other systems described above). For tenants actually paying the resulting rents, the effect of this is that the amounts paid by two tenants in identical properties run by landlords with identical efficiency can be completely different. The accident of where someone was allocated by a housing department can lead to them paying different amounts for the same thing. For most tenants, of course, gross rents do not matter at all for their net incomes (although they do affect their position in the unemployment and poverty traps). However, the government would like - for the kinds of reason described already and below - to move towards a Housing Benefit system which did not adjust fully for rent variations at the margin. If this happened without consistency in the principles determining social rents, the result would be to extend the unfairness of the existing anarchy in social rents to a larger and more vulnerable group. 15
Glennerster, Hills and Travers (forthcoming), chapters 10 and 11.
The lack of coherence of rent structures is not only an equity problem. It also means that there is no way for tenants or others to use the level of rents as a consistent measure of the relative efficiency of their landlord. An efficient landlord running a good service may be able to do so for costs a few pounds a week less than another. But this difference can be swamped by the variations which result from the different financial regimes. It should be noted, however, that some of the differences - such as the higher level of association than council rents - are not entirely accidental, but reflect deliberate government decisions taken with a view to minimising the public spending costs of subsidies and Housing Benefit. Higher rents mean lower subsidies and grant levels and so save public spending. But they also mean higher benefit spending, and draw more people into the benefit system. They also increase the RPI, with knock-on effects on spending on things like index-linked bonds and benefit levels. If they go high enough, these effects outweigh the subsidy saving. Association rents have less effect on the RPI than council ones, so the 'public spending minimising rent level' is higher.16 If this kind of accounting accident from the design of the RPI continues to dominate housing finance, there is little hope of a more coherent system emerging. As a corollary, however, moves towards 'convergence' are likely to have public spending costs. Polarisation of housing choices Paying for housing is not just about buying one thing. Table 1 shows how the different components of 'housing' are treated in the different sectors. Not only is subsidy available in different ways, but also people are largely constrained to opt for one package or another. Problems with this include:
The only way to 'get into the housing market' (for instance, because someone does want to buy eventually and is worried about future prices) is buying, almost always by buying a whole flat or house (apart from a small number of shared ownership schemes). The only way for, say, a poor elderly owner to get someone else to look after and organise repairs for them is by becoming a tenant. Some 'care and repair' schemes are available, but the owner would not be eligible for Housing Benefit on this part of their costs if they had a low income, in contrast to the position of tenants. It is hard for elderly owners to realise part of their housing equity to provide either a regular income or to defray care costs - 'home income plans' have not been very good value for money, partly because of the nature of the risks taken on by lenders in this situation and partly because of the interaction with other parts of the social security system. Conversely, it is very hard for social tenants to do what many owners can do on retirement, and move to a lower cost area, and even if they can negotiate a rare transfer, there is little gain from doing so, in contrast to the capital which owners can free up. The position of low income owner-occupiers As indicated in the introduction, it is difficult to consider policies towards social housing without touching on the treatment of the alternatives. Here it is important to note that as a result of the tax changes discussed above, owner-occupation is no longer positively l6 Also, the positive impact of reducing the capital grants to associations on the public finances is immediate, while the negative effects of rents via the RPI accumulate over time and, with a high discount rate, are weighted less heavily. By contrast, the trade-offs between recurrent council subsidies and benefit are immediate.
16
subsidised through the tax system in the way that it used to be, meaning that owners now meet the 'cost of capital' element of their housing costs in full (although, of course, many have still accumulated capital as a result of previous advantages). Even before Housing Benefit, low income owners are in this sense in a worse position than, say, most council tenants (although not necessarily than some high rent association tenants). The other advantages of owning clearly outweigh this disadvantage in most owners' minds. However, there is a number of public policy issues looming around marginal owner-occupation: Although people with mortgages can eventually receive some kind of help with mortgage costs if they end up on Income Support or means-tested Job-Seekers Allowance, there is now an extended gap before they do so. The policy assumption is that people will have savings or some kind of insurance policy which will carry them through. The work of Janet Ford and others shows that this assumption is misplaced, and our own work suggests that available insurance policies are for various reasons (some inherent to the nature of the risks involved) very poor value for money. l7 Arrears and (re)possessions continue at a significant level. If a recession struck, they could increase greatly. If this happens the public sector often ends up paying the bill either through Housing Benefit for expensive private renting or through the allocation of a social tenancy, with the permanent subsidy this involves. Neither of these forms of coping with mortgage default is very good value for money for the public sector (or good for the economy if the effect is to further depress house prices at that point). Quite apart from recession, there is a danger that borrowers have not adjusted to what appears to be the new world of low inflation. In the 1970s and 1980s, someone taking out a mortgage which was hard to service in the first few years could be fairly confident that inflation would mean a substantial amount of 'headroom' emerging later on as their income rose in cash terms but their mortgage costs did not. This gave them a margin of error to cope with later interest rate rises, temporary drops in income (e.g. one partner stopping work when there are young children) and so on. With low inflation, this headroom will be much smaller, which could leave borrowers more vulnerable to interest rises or personal income falls. Nor have borrowers necessarily woken up to the fact that paying off a mortgage as fast as possible may (in the absence of tax relief) now be the best investment open to them, leaving themselves vulnerable for longer than necessary. If an owner does get help with mortgage interest while out of work, this disappears entirely on return to work. This contrasts with the partial assistance which a low paid tenant may receive ftom Housing Benefit. This means that it is easier for owners to be 'better off out of work' than tenants - although it has to be said that there is not much evidence that this affects owners' behaviour, while many tenants are unaware or unconfident that they would still get any help from Housing Benefit in this situation.18 As explained above, low income tenants have their repair costs covered by Housing Benefit. The same is not true of owners. The state of the private rented sector The British private rented sector remains smaller and less effective than in comparable countries. This means that it is less effective in meeting housing needs than it might be - for instance, at particular stages of the life cycle or for those moving for work. This pushes some people into owner-occupation earlier and longer than might be sensible. It also increases the pressures on social housing. Some analysis suggests that the attenuated size of the British l7 l8
Ford and Kempson (1997); Burchardt and Hills (1997). Marsh and McKay (1993).
private rented sector is partly responsible for our unemployment levels.I9 Even if this is not entirely accepted, it is clear that a well-functioning private rented sector could help labour market mobility. But at the moment it is a relatively high cost option - the Joseph Rowntree Foundation private rents index suggests that landlords currently receive gross yields of 8-9 per cent (and over 1 1 per cent on flats) before allowing for any capital gains.20 This is not to say that the private rented sector would ever become very large - even with complete fiscal neutrality and rents which yielded rates of return comparable to those elsewhere, owner-occupation would remain a better option for most who can afford it. But it could play a much more effective niche role. Beyond competition from well-established owner-occupation (and from social renting) the barriers seem to be: The poor reputation of the sector. A history scarred by Rachmanism, past rent controls which made renting a poor investment apart from for the unscrupulous, and poor conditions have left it with a bad image for both potential tenants and investors. This may be receding. In most businesses one would expect the high returns available to attract in new investors, eventually driving rents down. This has only happened to a limited extent. A first problem is political risk: given our history of rent control, there is a fear that tenancies could suddenly be controlled, with lower rents combined with extended security of tenure, considerably reducing the capital value of an investment. This risk is probably also receding. However, a second risk may not be. A private landlord takes on the capital market risk of swings in house prices. The experience of the last 30 years shows that these can be substantial. Renting a property out for two years may yield a very good return - but there is the risk that property prices might again do what they did at the start of the 1990s. For long-term investors, short-term swings in capital values would be less of a problem but there is a conspicuous absence of long-term finance in residential letting (in contrast to commercial property), and initiatives like 'Housing Investment Trusts' have had no impact. Private landlords are unfavourably taxed not just by comparison with owner-occupiers, but also with other businesses. Depreciation - a major part of the long-run cost of running property - is not tax-deductible (although actual repairs are), while capital gains are taxed (in contrast to the routes for reducing or avoiding it now open to many other businesses). Even if these were changed, the tax position of landlords21would still be unfavourable by comparison with owner occupiers who pay no tax on their capital gains or 'imputed rents' (the value people receive in kind from living in their own home). Attempts to revive the sector in the face of these problems have proved expensive. The Business Expansion Scheme concessions where tax relief effectively paid half the capital cost of a new unit which was then rented out at a market rent for just a few years was, for instance, very poor value for money by comparison with, say, grants to housing associations which create long-term sub-market tenancies. The Housing Benefit system Last, but not least in thinking of how to create a more effective housing system, comes Housing Benefit. The advantage of Housing Benefit should not be forgotten: it means that no Oswald (1999). 1997 figures from Wilcox (1998), table 52. Apart from a limited amount of tax-free 'rent a room' by resident landlords, and of course, undeclared renting to lodgers, etc. l9
20
matter how low someone's income is, a tenancy will always be affordable (unless a private rent exceeds the 'local reference rent' or, for single people under 25, the 'single room rent'). Its drawbacks are, however, ~ e l l - k n o w n : ~ ~ Its rapid withdrawal (the 65 per cent 'taper') as income rises is an important contributor to the poverty and unemployment traps for tenants (and the savings trap for pensioners with small savings or small occupational pensions), including meaning that higher rent tenants can still be caught by it even when receiving the new Working Families Tax Credit. This is especially important for private tenants in the South, and for some housing association tenants. Making tapers shallower, however, always has the effect of widening the poverty trap, even as the trap is made shallower. Uncertainty about entitlement - and about how long it will take the councils running it to get the amount reassessed and paid - is one of the main reasons why tenants who are out of work fear that they will be worse off in work or will at least face a considerable risk in the transition to work. Take-up is less than 100 per cent - as a result some end up below what is supposed to be the state's minimum income level. It is more vulnerable to fraud than other parts of the social security system - not helped by the fact that gross rents do not matter (in the short run) to tenants receiving it. Automatic payment of rents to 100 per cent cases (and others on 'rent direct') is one of the reasons for reduced social landlord contact with their tenants. It saves rent collection costs, but means that landlords can be unaware of property or tenants' problems until late in the day - or may not even know that they have moved on. Even this system does not work smoothly, however. Housing associations and private landlords have to cope with arrears building up while, for no fault of the tenant, Housing Benefit is sorted out, or which emerge through breakdowns in administration. This makes it much harder to monitor and chase up arrears which are occurring for other reasons. All of this has been made much worse by the DSS's decision to pay Housing Benefit longer in arrears - all new tenants are pretty well automatically plunged into arrears. This makes it hard for landlords to monitor what is actually happening to their arrears - or for anyone else to tell whether their managements are controlling them effectively. In some parts of the country, the administration of recent measures taken to counter fraud - including more rigorous checks on identity documents - have further worsened delays in payment. For parts of the private sector, such administration problems are an important contributor to decisions not to let to HB cases.23 However, the key problem exercising policy-makers at present - and contributing to some of the above - is the effect of the way in which since 1988 all Housing Benefit has been based on 100 per cent of actual rent. Even for those who only receive partial HB, rent variations up or down are entirely offset by changes in benefit. Unlike any other commodity - including necessities like food - there is no advantage to a tenant from 'shopping around'. Within parts of the current system this may not matter so much. Social tenants are allocated to properties with little choice, so they have no option of shopping around anyway. Even in the private rented sector what evidence there is shows HB-recipients visiting as many potential properties as others. Housing Benefit for private tenancies is constrained by the 'local reference rent' system (and for those under 25 by the single room rent), with tenants paying 22
For recent discussion, see Kemp (1998,2000). The DSS has plans for administrative simplification, but implementation of these is awaiting decisions involved in the housing Green Paper.
23
the full amount of any excess themselves. For those expecting to receive benefit only for a short time, housing decisions may be little affected anyway. And lack of understanding of the details of the system may mean that people do not know enough to take advantage of it. But as part of a housing system which did have more choices and options for tenants or which was intended to increase productive eficiency pressures on landlords, the current system is bizarre. It also acts as something of a barrier to reforms which would cope with some of the problems outlined above. As part of a long-term reform of housing finance, there is much to be said for changing it. However, achieving that without causing hardship given our starting point is by no means easy. The next two Sections look at options for doing so within less and more radical options for the overall direction of reform.
4. Steps in the right direction Faced with this catalogue of problems, for some the solution is simple: get rid of all of the machinery of specific housing subsidies discussed above, including both social housing (by privatising it or transforming it into non-profit bodies competing in the market without subsidy) and Housing Benefit; let the market take care of allocating housing; and use more generous cash social security benefits and adjustments in the tax system (say, more generous income tax allowances and tax cred.its) to compensate for the losses to those with low incomes which the other measures would cause.24
It is important to be clear about the objections to this approach, even if some useful reforms would go partly in the same direction. Crucially, it would not achieve the core aims set out in Section 2 of supporting people with what would still be low incomes to afford a minimum standard of housing or of allowing a social mix in higher cost areas. Faced with high costs in particular areas, tenants would be left with the choices of moving elsewhere (to a lower cost region, if the flat rate aqjustments to social security were nationally set), reducing quality including through over-crowding, or making drastic cuts in other forms of consumption. Drastic area polarisation, new slums and substantial hardship are not attractive outcomes. To observe that current arrangements allow many people to make no trade-offs of this kind, and could be reformed to introduce them to some degree, does not imply that the only alternative is the full force of the market. But nor do these dangers mean that we should leave current systems as they are. These hndamental problems would remain even if reform was phased in slowly, but other problems could be partly dealt with through different kinds of phasing in: Given where we are starting from there would be very large gains and losses to particular households from changes of this kind. Even if a package of social security and tax changes meant that on average those with low incomes were compensated for other changes, there could be very large gains and losses in relations to individual households' incomes (even if benefits like Income Support varied on a regional basis in line with average rents). This could be ameliorated by a general improvement in benefits and so on, so that the average position was a gain, but only at a substantial net public spending cost (and still leaving some significant low income losers). It is all very well to tell people in what would become high cost housing that they could move somewhere cheaper to restore (or even increase) their net incomes, but such moves have substantial personal costs (and require affordable alternatives). Phasing in changes - or 'grand-fathering out' subsidies by limiting them to existing recipients - can help with the last problem, but care would have to be taken not to create disincentives to move or take a job, say, for fear of becoming 'a new case' as has happened with social security changes in the past. To avoid this, people would have to in some way 'take their old subsidy with them' as they moved. If flat rate allowances are paid to tenant householders to compensate for lost subsidies or Housing Benefit, there is a danger of creating an incentive for people to form new households - for instance, young people leaving home early, which would exacerbate the pressures of household formation. They could also be vulnerable to outright fraud (at the
24
For a recent proposal along these lines, see King (1999).
crudest where members of two households pretended to be one-another's tenants in order to create two additional 'householder premiums' via 'Giro drops'). Avoiding that problem suggests that compensating flat rate allowances should be paid to householders and non-householders alike (e.g. by simply raising benefit rates) - but this means clear gains to low income owner-occupiers and non-householders and higher net public spending costs (or inadequate compensation of tenant householders). This issue has to be faced even if a reformed Housing Benefit system is only partly based on a flat rate element. Of course, it could be argued that one of the anomalies of the current situation is that low income owners receive no help with their repair and maintenance costs, so a small flat rate addition to benefit rates or credits (up to £15-20 per week) could be justified as correcting for this.25 The disappearance of social housing subsidies which people currently retain as their incomes rise would inevitably worsen the unemployment, poverty and savings traps. In other words moves designed to help the housing market would harm the labour and savings markets. Again, changing the structure of things like tax rates and tax credit withdrawal could ameliorate the depth of the poverty trap - but only at the cost of extending it higher up the income distribution. There are clearly real problems here although it seems perverse that the housing market should be as comprehensively distorted as it is, if it is the labour market which is the real objective.
Neither unchecked market nor food parcel The combination of the fundamental aims described in Section 2 and these kinds of constraint suggest that the 'pure market plus cash redistribution' model should be rejected, even as a long-term aim. But at the same time enough problems have been outlined in Section 3 above to suggest that we should move away from where we are - particularly from the rationed 'food parcel' approach. This means moving towards systems which give those receiving housing support: more choice; more control and sense of ownership; more trade-off over costs, quality and location. If we don't do this we are facing the problem that a valuable national asset - the social housing stock - becomes the equivalent of so many black and white TVs in an age of colour, or even digital. It also means looking at subsidy and organisation from the point of view of housing consumers not of housing providers. In one sense the State needs to behave more as if it was the country's biggest tenant rather than its biggest landlord. This implies judging systems without any advance preference of provider, but rather by using the test of value for money. The sub-sections below suggest a series of modest steps in this kind of direction. The next section describes a more radical objective (although most of the steps below would be consistent with moves towards it).
Resident empowerment If we do remain within a world of social housing with sub-market rents, social tenants lack the power which consumers in other markets have of 'taking their business elsewhere'. The network of systems of tenant involvement and control which have developed over the last twenty years go some way to redressing the balance and should be maintained and extended. At the same time it is crucial to avoid putting residents in a position where they are set up to National accounts data suggest that in 1995 owner-occupiers as a whole spent about £14 per week on maintenance and other direct costs. The Family Expenditure Survey for 1997-98 shows average spending by owner occupiers on repairs and maintenance of £10.55 per week. The costs of periodic major repairs or depreciation could double this amount. 25
fail through taking on greater responsibilities than they want or can handle. Training and support can, of course, help but structures still need to be tuned to allowing residents to take a meaningful role in running their housing, if that is what they want. The most obvious issue here is scale. As suggested above, there may be a great difference between what is the efficient scale of a landlord in capital terms (as small organisations face high borrowing costs) and what is efficient in terms of day-to-day management and repairs, where groups of a few hundred units may offer not only a sufficiently large scale for efficient organisation of direct services but also clearer trade-offs and pay-offs to residents from local decisions. Estate-level budgets can offer both the efficiency advantages of a hard budget constraint on local landlord services and stronger options for resident involvement or control (as the experience of Estate Management Boards shows). Judging value for money in this kind of situation remains difficult, however, and doing something about it if it is poor even harder. Moves to a consistent structure of social rent setting as outlined below would help - it would be clear what underlying rent guidelines were supposed to be, and clearer that deviations from this said something about efficiency. Monitoring by the Housing Inspectorate and Housing Corporation would help point up where management is failing and could open up poor social landlords to a range of pressures and penalties (such as reductions in future credit approvals). There is, however, something perverse about saying that because a management has performed poorly, tenants should be further starved of resources, say for major repair. The more logical response is a mechanism for changing management. Some such power has existed - as exercised most spectacularly by the tenants of Walterton and Elgin Community Homes (WECH) when they used (since repealed) Conservative 'pick a landlord' legislation to secede from Westminster council. But this was a very involved process involving setting up an entirely new organisation and it taking on the full capital and revenue responsibilities of a landlord. As an easier option, if the financial system made it clear what rents and estate level spending ought to be according to nationally-set guidelines, one could envisage a more straightforward system under which residents could 'take their business elsewhere' to another social landlord on financial terms which would be neutral between the old and new landlords (or themselves if they opted to take the new organisation route).26 Such 'exit' powers, even if seldom exercised, have a strong disciplinary effect on providers. Three elements of organisation which would help resident empowerment are therefore: Organising and budgeting for day-to-day services locally, on a scale which may be much smaller than appropriate for landlord capital operations. Setting clear and consistent benchmarks for what day-to-day spending is expected to be, and how actual spending relates to rents across social housing, to help tenants judge eEciency (preferably with a range of landlords operating locally to allow comparisons to be made). Ensuring that ultimately tenants do have the power to 'exit' on level financial terms to another social landlord or to self-management.
26 Council tenants do already have the 'right to manage7under 1990s Conservative legislation, but this is also very cumbersome to implement.
Landlord financial structures The problems outlined in Section 3 suggest some of the following characteristics of a vision for social landlord financial structures: Separation of flows of central government cash intended for Housing Benefit and for supporting social landlords (already true for associations, and to extend to councils under resource accounting proposals, probably from 2002). Separation of local authorities' strategic role in ensuring that local housing needs are met from their role as landlords. This implies that what is currently council housing would be run by landlords at armslength from local authorities, eventually either as some form of local authority-owned or sponsored housing company or some form of housing association. Equivalent financial structures for all social landlords in terms of the relationship between rents and quality of accommodation, implying rent consistency between associations and councils (but not necessarily 'convergence' in average rents, as locations and quality vary between the sectors). Financial structures which allow adequate resources for landlords to plan and provide as they see fit for the long-term maintenance of their stock without continually having to come back to central government. This would also create stronger pressures on landlords to use those resources efficiently and to make the best judgement of the balance between day-to-day-repairs and longerterm investment programmes, and between borrowing and accumulating reserves (the position housing associations are already in). The logic of resource accounting supports the idea of providing properly for future major repairs and for ending the use of HRA 'surpluses' to pay for part of the cost of Housing Benefit. Instead such surpluses should flow into providing for major repairs and further tackling the repairs backlog - and indeed DETR is consulting on proposals that allow this to happen. Doing this, of course, has an immediate cash flow cost to government. However, it is important to look at this problem in balance sheet, rather than just cash flow terms. At the moment central government - the Treasury - faces a major outstanding liability: it will eventually have to give credit approvals and then pay for the borrowing costs on the spending needed to cope with the accumulated (and accumulating) backlog of council repairs. If it was prepared to crystallise this liability in the coming years and allow investment which would eat further into the backlog, central government would have a supply of carrots to encourage local authorities down this kind of route for running their housing. At the same time, it would reduce its long-term liabilities. The creation of the Housing Inspectorate and commitment to greater tenant involvement (e.g. Tenants Compacts) give it the mechanisms for deciding what speed a particular council should be allowed to move at towards greater autonomy as a landlord. Even if this decision was taken, immediate resources are inevitably limited, so it is hardly likely that the repair backlogs of all councils could be cleared at once. In addition, both the quality of councils as landlords and the repair needs of their stocks vary, so it would seem sensible to set up a series of stages down which councils could proceed, gaining greater financial autonomy as they went, but subject to approval in moving from stage to stage. Some could move very fast down this spectrum - indeed, the continuing programme of Large Scale Voluntary Transfers (LSVTs) means that some go 'all the way' at once. Others might move
very slowly. Ultimately to achieve the kind of vision described above, central government might have to become proactive in moving laggards through the stages, but by allowing those in best financial and management shape to move fastest, this kind of attention could be concentrated on those who most need support. The advantage of this kind of approach (rather than simply relying on LSVTs) is that it offers a mechanism for improvement across the council stock, rather than just in that which is transferred. The stages could take the following forms: 1.
2.
3.
4.
27
The use of 'surpluses' to pay for rent rebates would be ended for all authorities. All authorities would produce resource-based Housing Revenue Accounts, including calculation of an appropriate provision for future major repairs.27 The second element of the reform, which is already underway, is for a provision for newly accruing needs for major repairs - a 'Major Repairs Allowance' (MRA) - to be calculated for all authorities. Councils will be allowed to retain this allowance to use as they see fit, rather than relying on credit approvals for the newly accumulating need for major repairs. However, they will continue to seek credit approvals for their pre-existing repairs backlog. Retention of the MRA in this way could be made dependent on satisfying the DETR that they had produced a satisfactory business plan for long-term maintenance of the stock (including value-for-money in those plans), satisfactory reports from the Housing Inspectorate, appropriate tenant involvement structures, systems for decentralising budgets to estate-level, etc. If a council failed these conditions, it could revert to the current system of dependence on credit approvals for all of its major repair needs. Beyond this, however, the aim would be to move landlords out of the credit approval system altogether. To do this - to become 'stage 2' landlords - they would have to have cleared their major repairs backlog - either because it was small to start with, or because they were successful in persuading central government of the priority of their needs and quality of plans for dealing with them. These authorities would be autonomous (and on their own) in dealing with repair needs, but would still be within the annual system of subsidy balancing out costs like historic debt charges given their rent guidelines. 'Stage 3' landlords would be outside the annual subsidy system altogether. This would require their debt to be adjusted to a level which meant that the costs they were carrying for management and maintenance, major repairs provision, and debt servicing could be covered by rents at an appropriate level. This is in effect what the LSVT model currently achieves in one jump, but it could be approached more gradually through the stages described above. At this point the landlord would have full financial autonomy. Depending on how the borrowing of such a body was counted in the public accounts, it might have the ability to borrow without central control (as there would be no demands on or guarantee from central government resulting from such borrowing). As with LSVTs, progress to this stage could involve a size limit, and larger councils could be encouraged to split their stock into smaller management units as they progressed through earlier parts of the spectrum. One of the aims should be to have a range of social landlords operating within each area, rather than there just being one dominant landlord as there is now in many areas.
Under DETR proposals (DETR, 1999a,b), the latter will in effect &om April 2001, but taking out rebates from HRAs requires legislation and may not be in force until 2002.
Whether 'Stage 3' landlords are constituted as local authority-owned or sponsored housing companies or as independent housing associations (RSLs) is a separate decision, partly depending on public accounting arguments beyond the scope of this paper. The crucial aim here is sustainability of their finances and responsibility for their long run costs. Moving authorities down this spectrum would require continuing commitment of resources from central government to cope with the outstanding repairs backlog. By way of return, however, central government would be clearing its long-run liability to provide future credit approvals for those repairs, and eventually for providing housing subsidy. To achieve the vision described above, it would also be necessary to move towards a consistent set of rent guidelines across social housing, with common principles underlying the system of subsidy to local authority HRAs, grants to housing associations for new building, transfer terms under LSVTs (or for debt restructuring to reach 'Stage 37, and targets for consistency with existing housing association rents. Possibilities for doing this are described below.
Rents in the social sector On both efficiency and equity grounds it would make sense for financial structures for social landlords to converge on supporting a common framework of rents, with those rents bearing a consistent relationship to the quality and value of services and accommodation. This common structure could form the foundation of guideline rents in calculating annual subsidy to local authority landlords (if still at stages 1 and Z), of capital grants for associations, and of transfer terms for organisations moving between the two sectors. It could also be the basis of any rent caps imposed to control HB costs (depending on how far Housing Benefit was reformed to make these unnecessary). For the usual reasons, adjustment to this framework would have to be phased, but it would be best if this convergence made it clear what the ultimate target was and how it was c a l c ~ l a t e d .There ~ ~ are, essentially, three plausible models (all of which relate to the rent which would emerge for a fixed standard of efficiency and service, rather than the actual rent charged): Setting social rents as a percentage of what the market rent would be. There are two difficulties here: the absence of a market rented sector with properties of equivalent kind or location to base this on; and the degree of variation it would result in between properties. Some social rents would fall below recurrent costs, if the average was similar to today's. Setting rents in proportion to current capital values. These are much more easily established. However, even if the average stayed the same, this would result in even more variation in rents across the country than one would expect from the market - some components of housing costs are more of a flat-rate element than in proportion to capital values. This problem is why DETR switched away from rent guidelines which were based on capital values alone in the mid-1990s. A system based on achieving a consistent return on current capital values - what I have described elsewhere as 'target rents', designed to cover management and maintenance costs, plus depreciationlmajor repairs, plus a 'capital element', a percentage of capital The NHS offers lessons here, with the way in which convergence of resources for different health regions depending on their needs has been achieved through a clear target and an equal percentage reduction in 'Distance from Target' for all regions in any year - a far more transparent system than the DETR's 'damping' arrangements for converging on underlying rent guidelines and management and maintenance allowances. See Glennerster, Hills and Travers (forthcoming). 28
value.29 This is not, in fact, so different from the position of existing housing associations, receiving a percentage grant toward their capital costs, but no subsidy for other costs. The advantages of the last system include: giving a pattern of rent variation across the country and between dwellings in one area which would be comprehensibly 'market-like', but without the degree of variation one would expect in market rents; being based on factors which can be estimated straightforwardly (rebuilding costs to calculate depreciation1 major repairs costs; market values of untenanted property to give the capital element); giving rents which always exceed actual landlord (non-capital) costs; providing a clear way of assessing both capital grant requirements for new property and tenanted market value to a landlord as needed in transfer deals (as both depend on the sub-market return on capital); and consistency with the elements involved in the Government's general moves towards 'resource accounting' (i.e. allowing for assets in monitoring and planning public spending) as proposed for local authority housing.30
In such a system the crucial element which determines average and individual rents becomes the required (sub-market) net return on capital after other costs. To give an idea of the starting point, earlier rough calculations suggest that in 199516 for Great Britain as a whole, council rents after management and maintenance represented about 1.9 per cent of the capital value of the stock. Deducting depreciation or a major repairs allowance would reduce this to around 1 per cent, which would represent the 'capital element' in a system designed to keep average council rents as they were relative to trend capital values in 199516.~~ Subsequent real rent increases have increased this somewhat. Table 2 shows some purely illustrative calculations of the kinds of variation between different value properties which might result now with a capital element set at 1.1 per cent to give roughly the same national average as now. The table is intended simply to show how this kind of calculation would work, but it also illustrates the kinds of variation there would be between typical properties in different regions, and between the extremes. Unlike rents set directly in proportion to capital values the variation between regions would not be much different from now. However, this kind of approach would give a clearer signal than existing DETR 'guideline rents' on relativities between individual properties, with a consistent logic across the country. The table also shows the rent variations which would result from a much higher capital element; this is returned to below. The level of existing housing association rents would generally imply a rather higher 'capital element', so consistency between the two sectors would either mean higher council rents or lower association rents, or a combination of the two. The overall system is, however, consistent with movement in either direction. 29
Hills (1988). A key issue discussed there but not pursued further here is the need to use some measure of trend capital values, rather than rents moving rapidly up and sometimes down with the house price cycle. 30 Indeed, a system of this kind would mean that an authority's rent guideline could be derived f?om within its resource accounts, rather than set externally. 31 This does not, however, measure the full net yield which landlords are receiving from rents calculated in this way if it is anticipated that capital values will rise in real terms in the long run, as they have historically. For instance if the capital values on which these calculations are based rise at a trend rate of 'RPI plus 1', the present value to the landlord of the resultant rising net rental stream of 1 per cent of capital value is equivalent to a net yield of 1.4 per cent: (with a real discount rate of 3.5 per cent). It is this which measures the sub-market return and hence calculation of subsidy in economic terms. For further discussion and derivation of figures, see Hills (1998), pp. 177-9.
The system could also be run on the basis of varying capital elements across the country, that is with an explicit decision that, say, the percentage capital subsidy should be higher in high cost areas. While this would be rather less neat as an overall framework, it could be consistent with the key aim of channelling support where it was most needed to allow people to afford reasonable housing. The same objective could be met directly (and more smoothly) by varying the size of the capital element for properties with above average values. This is returned to in Section 6.
Housing Benefit If social rent relativities did bear a more consistent relationship to the value and quality of accommodation, one of the objections to moving away ftom Housing Benefit based on 100 per cent of rent would be removed. However two constraints would remain: first, the fundamental aim of actively encouraging housing consumption by the poorest, which argues against moving towards a completely flat rate system, even in the long run; and second, the potential scale of losses for households which would find it difficult to move from more expensive accommodation than they would have chosen if they had known a flat rate element was being introduced.
In terms of modest moves towards re-introducing 'shopping incentives' into the HI3 system, four options suggests themselves: At the most modest end, we could move away ftom the present principle of (in theory, at least) adjusting Housing Benefit week by week to changes in circumstances, such as in rent or income. For instance, the new Working Families Tax Credit (like its predecessor, Family Credit) is set for six months at a time. In some other European countries, housing allowances are set annually on the basis of the previous year's income. Moving away ftom immediate adjustment of benefit would allow, for instance, a temporary gain ftom a move to cheaper accommodation and would lessen the immediate effects of the poverty trap.32 Some change of this kind would have to be brought in anyway if Housing Benefit for working families was integrated with the WFTC in some way. Fixing the amount of benefit for a longer period would have administrative advantages and might also help some of the uncertainty problems in the transition off benefits and into work. This could in principle be done either symmetrically, so that no change in circumstances affected benefit until it was reviewed, or asymmetrically so that people could request a reassessment if their situation deteriorated and HI3 entitlement rose. The most obvious way to introduce 'mild' shopping incentives would be to move towards a hybrid HB system based on a high percentage of actual rent and a small flat rate element reflecting a low percentage of average rents (for that location and tenure). Starting with a low percentage would control the immediate gains and losses, but over time this could gradually be increased (over say 10 years) to the long term result desired. An objection to this is that a low percentage would have only minimal effects in some parts of the country (leaving the 20 per cent minimum Poll Tax payment problem of collecting very small net amounts from tenants at some administrative cost). A high percentage would, however, mean significant individual gains and losses in high cost areas. An better alternative might be to set the flat rate amount in cash and adjust the percentage contribution to actual rents depending on local average rents for the tenure. Thus for 32
See DfEE (1999), chapter 6, for a suggestion of this kind, building on existing 'benefit run-on' arrangements.
instance, if the flat rate element was £10 per week (possibly paid via additions to benefit and tax credit rates rather than through 'Housing Benefit'), the rest of HE3 would be based on 80 per cent of actual rent if the local average was £50, but on 90 per cent of actual rent if the relevant average was £100. As well as administrative advantages and control of large individual losses33,this would mean that some parts of the country could move quite quicltly to a system with more normal incentives for housing choices, while retaining high sensitivity of benefits to actual costs in the parts of the country with the greatest affordability problems (again consistent with the key objective^).^^ Finally, one could argue that the 'shopping incentive' issue is most relevant for younger and better off tenants only - in effect for those in work. This was the motivation behind earlier proposals for a 'dual taper' Housing Benefit system, under which low income tenants would have benefit calculated as now on the basis of 100 per cent of actual rent and a high rate of withdrawal with rising income, but higher income ones would have it based on a lower percentage of actual rent but a slower rate of withdrawal (taper).35 The actual payment would be based on whichever was more generous. This would have a net cost to the Government. However, it would simultaneously improve incentives to work, give shopping incentives to those for whom they were most relevant, but protect those with the lowest incomes. It could also be used as a basis for integrating Housing Benefit for working families into the WFTC system, if say, the upper taper was used for tax credit cases (and with the housing credit presumably fixed for a period with the effects described above), although this would require careful design to avoid awkward jolts between the two parts of the system. Note that in the middle two options for gradually transforming HI3 into a 'hybrid' system the rules affecting the social and private rented sectors would become different in some respects (such as the size of the local flat rate element or percentage of actual rent used). This is not particularly desirable as a long run characteristic of the system, but would be necessary so long as social rents were sub-market and big losses to private tenants (or high public spending costs from 'over-compensating' social tenants) were to be avoided. The treatment of low income tenants and low income owners would also remain different. Proposals for re-introducing shopping incentives into Housing Benefit and for producing more consistent rent structures go together. A third element - how to implement which goes beyond the scope of this paper - would be to change allocations systems in a way which would allow a greater role for social tenants to have more choice over where they were allocated. What is important, however, is that a corollary of introducing some choice (albeit limited) is that gross rents vary in some consistent way with quality and value, and that there are some net effects of different choices (however modest). A final observation is, however, that there is probably little point in introducing very modest 'shopping incentives' if they are not seen as a first step to something more significant. What is done in the short-term therefore depends on longer-term objectives. Fuzzy tenures: Alternative ways of supporting housing costs One of the problems with the current position discussed in Section 3 is that the main instruments we use to support housing costs are very 'lumpy' - we allocate a socially-owned Or cost of compensating for them if, say, existing recipients, were protected against such losses. This would have an analogous effect to the process by which German private rents were decontrolled, leaving the cities with the greatest resultant effects until last. 35 For more detailed discussion, see Hemming and Hills (1983) and Hills (1991), pp.176-9. 33
34
dwelling with a substantial subsidy for life or not at all; we subsidise rents at 100 per cent at the margin or not all. This is not necessarily an efficient way of meeting varying life-cycle needs. There are some limited schemes which break this down a little, mainly through housing association low cost home ownership schemes: 'equity-sharing' for new developments (where people buy, say, half and rent the other half at a subsidised rent); and 'do-it-yourself shared ownership' (DIYSO) where people can find their own existing property on the market and set up the same arrangement. Some social landlords are also developing property to let at market rents. These schemes are relevant to a fairly narrow income band and are currently less favoured than they were by the previous government, but they embody principles which could actually be extended as part of a strategy to widen housing choices and to create neighbourhoods with a greater social mix. Such a strategy could include: Continuation of mixed tenure with low cost home-ownership components in new developments. Advertising and renting out some properties in what are currently solidly council housing areas at market rents or cost-covering rents (that is returning an unsubsidised return on capital). These would have limited security of tenure and no Right to Buy at discount. This could help mix existing areas socially, and would generate income towards major repairs backlogs or purchases in other less-predominantly rented areas. Equity-sharing offers for existing properties (with the right to 'staircase up' but not the Right to Buy at discount) rather than just in new mixed tenure developments. This could be another way of attracting moderate income households into predominantly rented areas, with what is effectively partial subsidy until full purchase (staircasing to 100 per cent) happens. This could help, for instance, a young couple through a low income period while they have small children, but does not lock the state into heavy permanent subsidy. In the other direction, as understanding of equity-sharing spread, more opportunities could be created for people to 'staircase down', by selling (to a social landlord or someone else) a share in the equity of their property. This could help, for instance, unemployed owner-occupiers who might currently have to sell and lose their property altogether (it does not solve negative equity problems). It could also help older owneroccupiers who wanted to realise equity to meet care costs, major repair costs or to boost annual incomes. The combination of measures like these and the regime suggested above for setting target rents would create a major new opportunity to attack the current polarisation of housing choices fiom the other direction. Through equity sharing, social landlords would have future assets (the amount they receive when equity shares are bought out, and, potentially, the amount they are paid until then) linked to local house prices. Similarly, their subsidised properties would generate net rents which were a proportion of future capital values, again linked to local house prices. It would make sense for such landlords to have liabilities which had the same profile. In other words, instead of borrowing in cash , they could raise money with their liability linked to what happened to local asset and rent levels. In turn, those supplying this funding could bundle it together and issue 'regional housing bonds' - creating the way which is currently missing for people to 'get into the housing market' in a particular area, without having to become a full owner-occupier there (with potential benefits for delaying premature home-ownership amongst other advantages).
At the same time, more flexibility over allocations and stronger regimes for tenant control and participation would give tenants within the subsidised (and still rationed) sector more of the control which owners have directly and which market renters have through the ability to move elsewhere. In all these ways, the current sharp contrasts between the positions of tenants and owners could be blurred.36
36 Another possibility might be opening up the payments for 'care and repair' schemes to HB eligibility. However, if owners were already net gainers from some general adjustment of benefit levels as HB was reformed, this would be inappropriate (they would effectively be getting such help already).
5. A more radical objective Few of the steps outlined in the previous Section are of themselves major breaks with the current system. However, they would take us towards a system which would be an improvement on where we are now in several important respects: They would establish a system under which all social landlords were supported in equivalent ways, with coherence in the structures of subsidy they were receiving, and internal financial structures which were consistent with efficient husbandry of the assets they own. This could be part of an overall 'reinventing' of social housing - arguably an essential one. Social housing rents would remain sub-market, but clearly related to quality and value of accommodation occupied and landlord efficiency. More extensive devolution of revenue budgets to smaller scale units would increase opportunities for resident involvement and control, and would make landlord efficiency easier to gauge. Ultimately, a coherent financial structure could also give tenants a more credible threat of taking their housing and business elsewhere if managements failed to deliver. Limited Housing Benefit reform would introduce some trade-offs to be made by a greater proportion of tenants between value of accommodation and its rent. This might lead to some more efficiency in housing choices. With gross rents having at least a partial effect on more tenants, pressure on landlords would increase to strike an appropriate trade-off between rents and spending without such heavy reliance on existing inflexible and anomalous rent caps. Coherence in rent structures and 'mild' shopping incentives would be corollaries of allocation systems which gave a greater role for tenants having somewhat more choice in their initial allocation, and might free up some transfer opportunities. Market-renting and equity-sharing opportunities for some of the existing council stock could help reduce area polarisation, and give housing options for people with moderate incomes without granting - in effect - a potentially life-long subsidy. A wider range of equity sharing arrangements, and a market in regional 'housing bonds' would help blur the current sharp choices between tenures, as would possibilities for some owners to realise equity by 'staircasing down' to avoid repossession, cope with care or major repair costs, or to generate income in old age. Such advantages would be worth having. There are, however, limits to how far they could be pushed. In particular, Housing Benefit reform is heavily constrained by the need to protect existing tenants in relatively expensive accommodation. This means that any 'shopping incentives' are likely to be fairly modest. Also, social renting - with rents well below market levels - would still have to be rationed. This means only a very limited degree of choice can be introduced, and reduces tenant power in relation to their landlords. The large differences between social and private rents stand in the way of a common flat rate housing element in benefit. Retaining an extensive system of Housing Benefit with amounts depending on actual rents even for comparatively better-off recipients complicates administrative moves towards integration with a tax credit system for those in work. In short the kind of measures outlined help blunt some of the problems in the housing market, but could hardly be claimed to remove them altogether.
It is therefore worth exploring what would have to be done to make a bigger difference, and whether some of the measures suggested above could be pushed further in a way which would help loosen up the constraints in other directions. In particular, suppose that it was taken as objectives that: 1. Over a period of years, social landlords should set rents which covered their long-run costs including an economic return on current capital values, in effect moving to a regime of zero economic subsidy. 2. Housing Benefit should be structured so that it incorporated significant 'shopping incentives', for instance, through a flat rate element of, say, £20 per week for a typical household at today's prices (with locally varying percentages of actual rent used to calculate the rest of benefit).37 Economic cost-covering social rents were the aim, for instance, of the proposals of the 1985 Inquiry into British Housing (chaired by the Duke of Edinburgh). Those detailed proposals greatly underestimated the rent increases which would then have been required, but the advantages they were looking for remain. If social landlords were effectively paying their way, it would have two advantages. First, one could escape from a lot of the current paraphernalia of rationing and introduce greater competition between providers in a situation where tenants would have a much more realistic 'exit' power. Second, a common flat rate element could be used in a reformed Housing Benefit across the rented sector, rather than different rules applying to social and private sectors. In effect, this represents a partial solution to Beveridge's 'problem of rent'. Social landlords would be operating as partly competitive non-profit bodies covering their long-run costs, rather than as rationers of subsidised goods. The aim would be to direct housing-related subsidy to individuals, not to particular dwellings. Previous attempts to do this have ended up piling more and more onto means-tested Housing Benefit, with worsened incentives, poverty traps and so on . Is there an alternative? And how could existing tenants who are not receiving Housing Benefit be compensated for or protected against the significant rent increases which a move to economic rents would entail? One answer might lie in the equity-sharing ideas discussed in the previous section:
3. Existing social tenants would be given an equity share in the property they occupy, reflecting the value of their current expectation of paying sub-economic rents. The landlord would 'pay' them a return to these equity shares which would bring the full economic rent back down to current levels. For a tenant staying put, this would have a neutral effect. In financial terms this would not, in fact, be such a major change for landlords either. Suppose one was starting from a position where the rent covered management, maintenance, major repairsldepreciation, and a balance equivalent to something over 1 per cent of the capital value of the dwelling (not far from the current position for average council rents). Suppose further that the real (after inflation) cost of capital is 3.5 per cent, and that rents and capital values are expected to continue to rise at 1 37 SOthat in an area where average rents for the relevant kind of accommodation were £50 per week, Housing Benefit would be based on 60 per cent of actual rents (plus the £20). In an area where they averaged £80, benefit would be based on 75 per cent of actual rent (plus the £20). A flat rate element of £20 is suggested as it would be preferable (but expensive) for this to be built into general benefit rates to avoid the incentives set up by 'householder' or 'tenancy' additions. This would mean extension to owner-occupiers as well, and £20 is perhaps the maximum amount one could justify as reflecting their long-run repair and maintenance costs.
per cent above inflation. This means that an economic rent would require a 'capital element' of 2.5 per cent38in addition to covering the other costs (the results of this are illustrated in Table 2 above). The current rent is yielding less than half this - if the tenant remained in perpetuity, their financial position is already that of someone owning more than half the equity in their property. This is reinforced by succession rights and the Right to Buy someone who can exercise the Right to Buy at a discount of 50 per cent also has potential financial rights equivalent to a half equity share.39
In other respects the change could be profoundly radical. In effect, it would be making explicit what are currently implicit rights. Tenant rights of involvement and control could flow clearly from an explicit stake in ownership (it could even be called 'stakeholder housing'). It could open up the way for people to arrange much easier transfers by trading in their equity shares (and then paying the balance of the economic rent, higher or lower, in the new property in the same way as anyone else). As the country has a very well-established system for trading wholly-owned property, there does not seem any problem with this in principle. Granting such equity shares to existing tenants4' - the 'Right to Own' - could replace Right to Buy discounts. As with the kind of equity-sharing proposed in Section 3, people could have the ability to 'staircase up' by increasing their equity share. The strength of this proposal is that it would protect existing social tenants from the distributional effects of raising social rents to economic levels, allowing subsidies to be 'grand-fathered out', but without locking them into their existing property. It also means that worsening poverty and unemployment trap problems would be avoided for this group.41 It would remove one of the major objections to moving social housing much closer to the market. But what of new tenants? For one group, as with the proposals in the previous section, unsubsidised tenancies would be allocated in a way not very different from a commercial landlord and full economic rents would be charged. Social landlords would no longer be as constrained by grant allocations in creating new developments. Access would no longer generally be on a rationed basis, and movement would be a great deal easier, although there 38
This is contrasts with the Treasury's calculations in 'resource accounting' that an unsubsidised return on capital should be 6 per cent. First, that ignores the return to a landlord through real capital gains (or, equivalently a rising real rental stream). Second, 6 per cent is far higher than market real interest rates or the cost of capital to owner-occupiers. In effect, it incorporates the very large risk premium applied by the Treasury in assessing public investments. It is not obvious that this is appropriate here. 39 DETR's own calculations of the 'adjustment factor' needed to convert vacant possession values into 'existing use values', given current council rents, suggest figures ranging fi-om 48 per cent in London to 64 per cent in the North East (DETR 1999b, p.23). In other words they suggest the capitalised value of paying sub-market rents for tenants ranges from a third to a half of the property's value. 40 Calculating the equity share might involve a minimum period of residence and rent payment - say ten years. For existing tenants of shorter standing at the time of the change their right to trade in their equity share by moving could build up over the period until this had been reached (in a similar way to Right to Buy discounts). Periods for which rent had not been paid because of arrears presumably would not count towards qualification for the equity shares. 41 There is a rather important technicality in calculating Housing Benefit under the reformed kind of system assumed here, with a flat rate element and using only a percentage of actual rent: for this group of tenants, benefit should be calculated on the full economic rent and then the discount for being an equity sharer deducted from the entitlement. If benefit was calculated on the basis of the net rent charged to equity sharers, they would gain unduly fi-om the flat rate element, despite having been protected fi-om the higher rents. Similarly there might have to be restrictions against selling out an equity share for cash, using the proceeds for something else, and then claiming full Housing Benefit.
would be a continuing need for a needs-based entry route (even if rents not subsidised in economic terms) to maintain a social mix in high demand areas.42 Some tenants would, of course, subsequently require Housing Benefit. For those whose incomes dropped substantially, this would cover the whole of an average local rent (varying most with the actual rent in high cost areas, but not by 100 per cent as now). For new entrants on this basis, there would be the opportunity to purchase equity shares if they wanted, and to 'staircase up'. This first group of new tenants would include some who would not have moved into social housing under current arrangements (possible through more open access and the development opportunities opened up) as well as some of those who would previously have moved into subsidised social tenancies, but who have reasonable income prospects (one hopes an expanding group with rising affluence and falling inequality). The issue is what happens to new entrants who do not have reasonable income prospects. There are two logical options. One is to charge the same level of rent to all and rely on the Housing Benefit system. This would not be so different from now (except where for some reason a particular property's rent was a long way from the local average - which ought to be rarer than now given the rent setting described). A downside of this would be that people would face longer poverty and unemployment traps than they would be with current rents particularly in high cost areas, as the rent levels in the last row of Table 2 suggest. Increasing the highest council rents in London by £50 per week and hoping that the benefit system would take the strain would have serious work incentive effects. A second objection is that this would create big differences in rights between those who were tenants at the time of the change - who would be granted equity shares - and those who became tenants shortly afterwards, who would not. The second option would be to have a selected entry group of new tenants who would be granted or would build up a free equity share, which would reduce their costs in the same way as for existing tenants at the time of the change. This would cut back the labour market disincentives. It would require continued substantial capital subsidy (or long-run revenue underwriting) for social landlords. It would mean continuing a selection between full rent tenants, and those receiving a rationed supply of valuable long run subsidy (although not tied as much as now to a particular dwelling). In housing market and management terms, this would be a less desirable option - in effect proving 'need' would carry a more explicit prize than now, and a very valuable one in some parts of the country.
42
Preserving such a route would imply a consequent need for protection against full staircasing up where this would erode the flow of needs-based entry - as some in rural areas, and for part of stock in all higher value areas.
6. A synthesis: back to the objectives The motivation behind such a radical long term objective would be the transformation of the role of social landlords which it would allow. It would open up a large part of the housing stock which is currently completely outside them to far more normal market pressures while protecting existing tenants, allowing far greater opportunities for mobility. It really would mean the end of the 'food parcel' approach to social housing, while making a transfer of explicit assets to all existing tenants. It would open up the opportunity for far-reaching Housing Benefit reform - while retaining high level of subsidy on actual housing costs in high demand areas. Where it falls down is in the treatment of new social tenants in high cost areas. With much higher rents in such areas, both the increased Housing Benefit route and the alternative of the granting equity shares to a selected sub-group of new tenants raise problems. The question arises of whether there is an intermediate route between the modest rationalisation discussed in Section 4 and the radical transformation of Section 5. At this point it is worth returning to the main justification set out in Section 2 for the combination of sub-market rents and housing benefits whose size is directly related to actual costs. The key issue is affordability of 'decent' housing by those with low incomes. We do not want to leave things entirely to the market because of the dangers of very low quality housing and extreme area polarisation which would result. We do not want to load everything onto the Housing Benefit system because in some areas this would cause very deep unemployment and poverty traps. Flat rate allowances as a replacement for even a large part of Housing Benefit in high cost areas would create hardship for some - Beveridge's 'problem of rent'. But these problems have one thing in common: they all relate to the difficulties in high cost areas and regions. They do not rule out a system which has lower levels of economic subsidy and greater relative reliance on a flat rate element within housing benefits in low and medium cost areas. As far as two of the key decisions are concerned, this suggests the following long run objectives:
1. Within the proposed 'target rent' system - where guideline rents for council housing and capital subsidy for housing associations are based on achieving a set return on current capital values (after other costs) - the required return could vary depending on the value of the property. This could be done using regional variations, but this both creates discontinuities at regional borders and still leaves a problem within high cost parts of low cost regions. A better solution would be to set a 'low subsidy' required return up to the national average capital value of social housing (for example, following the illustrative calculations above, 2.5 per cent up to a current value of &44,000), but a return implying a higher level of subsidy for the excess of values above that (e.g. the present 1.1 per cent or so net return within council housing on the part of capital values above £44,000).~~ 2. Incorporating a nationalflat rate housing costs element into benefit rates and tax credit structures, with locally varying percentages of rent covered by explicit Housing Benefit. For the reasons given, it would be better if the flat rate element was a standard part of benefit rates for all tenures, and not given as a 'householder' addition (it would therefore vary with household size in the same way as the rest of benefits). There is an equity 43
Thus the capital element for houses worth up to £44,000 would be 2.5 per cent of value. For a property worth £64,000, the capital element would be 2.5 per cent of £44,000 plus 1.l per cent of £20,000 (i.e. £1320 per year).
justification for this going to low income owner-occupiers up to a reasonable level for their repair and maintenance costs (including periodic major repairs), but it would seem hard to justifjr going beyond £15-20 per week to an average household at current prices on these grounds even if the public spending cost of increasing benefits for this (low income group) was accepted.
In both cases it would be sensible to approach these changes in stages, not just because this allows adjustment on the part of tenants, but because it also allows experience with the effects of the changes to build up - and if there are problems to adjust what is being done. Thus, the move towards higher net returns on the first part of capital values could be phased in - say over 9 years by raising the threshold in annual steps of £ 5 0 0 0 . ~ ~Starting in the lowest cost parts of the country, social housing would then gradually move onto a low subsidy regime until the 'rising tide' reached areas with average capital values. Similarly, the national flat rate housing costs element in benefits could be built up over time - perhaps starting with an initial amount that would be meaningful and then in smaller steps until the target was reached45 If we were starting from scratch, the end result of this would be a logical solution to the fundamental objectives of policy and the constraints under which it operates. General subsidy to social tenants (in economic terms) would be modest in low and medium cost areas, allowing much more fluidity in the housing system and more realistic opportunities for choice and mobility. But subsidy would remain high in areas (including higher cost parts of cities in low cost regions) where it was needed to allow a social mix and to avoid severe labour market problems. Housing Benefit would no longer be based on 100 per cent of rent in any region, so there would be 'shopping incentives' across the country - the corollary of more choice. However, even in the lowest cost regions, low income households would still be receiving significant specific assistance with their actual costs. In the example given above, if the flat rate element had reached £20 per week, and the relevant average local rent was £50, the variable part of Housing Benefit would be based on 60 per cent of actual rent. If local rents were £80, the variable part would be based on 75 per cent of actual rent. Someone paying the local average rent could be treated as they now are, but they would face (modest) costs or gains for rents higher or lower than this. Again, higher subsidy at the margin on actual housing costs would be concentrated on the areas where affordability problems were greatest. As a further corollary of such reforms - and the introduction of significant shopping incentives everywhere - the need for the crude limits set by local reference rents, the single room rent, and the rent caps for councils and associations would be reduced, if not eliminated. It would be better to end up with a system that incorporates some incentives for all rent levels, than none for most and then very sharp ones for those above particular thresholds. This would be the equivalent of an annual increase in guideline rents for local authorities of£ 1.34 over and above the effects of inflation and rising capital values (if the starting capital element was indeed 1.1 per cent). For housing associations existing rent levels may well already be close to those suggested by the long-run target already. 45 For instance, if the target was £20 at today's prices, this could be achieved by an initial amount of £4 and increasing in £2 steps for an average-sized household, with appropriate adjustments to the percentage of rent Housing Benefit was based on.
44
Finally, however, it should be noted that these two moves have different distributional effects. The Housing Benefit reform could be introduced with no net distributional effects the average case in each area could be left unaffected (and the national flat rate element proposal keeps the scale of gains and losses around this tightly constrained). But an increase in the capital element of target rents in the way suggested could mean a rise in council rents building up to a maximum of £10-12 per week at the end of the transition period (over and above the effects of inflation and rising costs). This would mean real losses for existing tenants who were not receiving Housing Benefit. There are three approaches which could be taken: To argue that this would only accrue gradually and was the price of a move towards a more flexible housing system, accepting the adverse distributional effects. Existing housing association tenants are paying rents at this kind of level already, so this amounts to the phasing out of a relative advantage. Protect existing tenants by allowing them to retain the right to a lower level of rent as long as they remaining in the tenancy. This 'grandfathering out' is effectively what has happened to housing association tenants since 1989, with a rent gap between 'secure' and 'assured' tenancies. Eventually the gap could be closed after a long transition period (again as is now happening to housing associations). The downsides of this are the creation of another discontinuity within rent structures and the danger of barriers to mobility if the protection does not 'follow the tenant'. An alternative would be to use the 'equity share' ideas described in Section 5 at a more modest scale to protect existing tenants: as rents rose, they would be granted an equity share which would be increased over time (and with it the discount from the general level of rents).46
46 An
equity share giving a discount to offset a rise in rents of £10, with a capital element of 2.5 per cent would for instance be worth about £20,000. The value of the equity share would be less in properties with below average values where the rent increase was smaller (e.g. £10,000 for a property worth £22,000). The proposed long-run structure would mean that there would not need to be any greater value of this 'rent increasecompensating' equity share than the £20,000 figure in higher value parts of the country This limit could be seen as desirable in regional equity terms.
7. Conclusions The summary at the start of the paper brings together some of the options which could be pursued to rationalise existing arrangements regardless of the decisions taken on the long-term objectives for the degree of economic subsidy within social housing and the structure of Housing Benefit. These recommendations would help create more mixed communities, greater tenant voice and influence, long-term viability of social landlords, coherence in social rent structures, and less polarisation in housing choices. Section 6 sets out a synthesis of some of the choices for the key decisions on rent levels and Housing Benefit which recognises that the problems we are dealing with are different in high and low cost areas. The objective is to support those who would otherwise not be able to afford an acceptable standard of housing while avoiding social polarisation. To achieve this, the degree of specific housing subsidy (whether through below market social rents or the degree to which Housing Benefit relates to actual individual rents) needs to be much higher in high cost areas (including those within low cost regions). In low to medium cost areas there is much more scope to use general measures of redistribution and less reliance on specific housing subsidy, allowing an escape within such areas fkom many of the problems resulting fkom rationing. Table 3 summarises how all these elements would fit together. Running through the paper is a theme that we are still prisoners of decisions taken in the 1940s. Beveridge failed to integrate housing costs into his otherwise comprehensive plan for social security because of what he described as the 'the problem of rent' - housing costs which varied widely between and within regions in a way which could not be dealt with by flat rate allowance^.^^ In contemporary discussions we are no further on - perhaps because we are always, like Beveridge, looking for comprehensive solutions rather than identifying a series of partial ones. Decisions taken in the 1920s and immediately after the second world war established council landlords as the main route (apart fiom assistance h m benefits) for supporting the housing of those with low incomes. This left us with a landlord structure very different fkom those of most of our European neighbo~rs.~~ These decisions may well have been appropriate solutions to the housing problems of fifty years ago. Indeed, in pointing to contemporary problems it is important not to forget the huge improvement in housing conditions to which they have contributed. But that very success means that the nature of the problems has changed, and with them appropriate solutions. We are no longer dealing with gross housing shortages across the country or widespread overcrowding or lack of basic amenities. We are also dealing with a much more affluent society, albeit a still unequal one. Furthermore, any wideranging reform of housing finance is almost bound to take effect over a long period - adjusting fiom current arrangements would otherwise be too painful. In setting objectives we therefore have to look at the needs of Britain in 2010, rather than just focussing on 2000. This need to look at reforms which are geared to a long-term goal has a political downside: the benefits will also arrive in the long term while some of the upheavals required will come more quickly. But if we never make the first moves towards a system better attuned to contemporary needs we shall forever be giving the answer that to get there we should not be starting from here.
47
48
Beveridge ( 1 942), particularly pp.77-84. Power (1993, 1997).
Table 1
Elements of housing costs by tenure Owner occupation
Management
Self (some elements not required)
Maintenance1 repair
Purchase / DIY / neglect
Major repairs1 depreciation
Self (pay or lose value) Possible improvement grant
Cost of capital
Pay (implicit or explicit)
1
Private renting
Housing rent1 Benefit
Local authority
f
rent Credit Approvals
]
In rent 1
Part in rent / part Part from rent / in subsidy part fiom capital grant
Purchase of asset Purchase / take housing investment market risk
Risk / gain to landlord
Care
May be provided / by landlord
Separate from housing
Housing association
-
/
Table 2 lllustrative calculations of 'target rents' England
North
London
'Low North'
'High London'
Capital value (£) Management and maintenance (£1year) Major repairs provision1 (£/year) Capital element @ 1.1% (per year) Rent loses (voids, etc at 4%)
Total (Slyear) Total (Elweek) Actual 1999 (approx) Rents with capital element @ 2.5%
'
At 0.9% of capital value nationally, 1% in North (with same amount in 'low north'), 0.8% in London and 0.75% in 'High London' (allowing for varying land element in capital values).
Table 3 Summary of potential reforms Shortlmedium term
Long-term (10 years)
Social landlord finances
More local budgeting for day-to-day services. Financial structures of all kinds of social landlord LAs move towards separated finances for housing on same basis. with long-term viability: retain HRAs; catch up on major repairs; debt adjusted to end recurrent subsidy.
Social rent structures
Consistent relationship between rents and services across social housing. Rents set to cover: management and maintenance; major repairs; percentage of capital value.
Raise capital element towards lower economic subsidy up to national average value; retain lower percentage for excess over this (i.e. concentrate economic subsidy in high cost areas.) Protect existing tenants through granting equity shares?
Housing Benefit
Modest moves to shopping incentives, e.g. fixing benefit for a period. First stages in long-term reform towards a system with a flat rate element.
Move in stages to significant national flat rate element in benefitltax credit rates, with locally varying percentage of actual rent used to calculate benefit (higher in high cost areas).
Fuzzy tenures
Continuation of mixed tenure in new developments. Market renting of some social stock. Equity sharing offers for existing social stock. Housing bonds linked to property values.
References Beveridge, W.H. (1942), Social Insurance and Allied Services, Cmnd. 6404. London: HMSO. Burchardt, T. and Hills, J. (1997), Private welfare insurance and social security: Pushing the boundaries. York: Joseph Rowntree Foundation. Department of the Environment, Transport and the Regions [DETR] (1998), A New Financial Framework for Local Authority Housing: Resource accounting in the Housing Revenue Account. London: DETR. DETR (1999a), Calculation of the Major Repairs Allowance: A consultation paper. London: DETR. DETR (1999b), Guidance on Stock Valuation for Resource Accounting: A consultation paper. London: DETR. Department for Education and Employment [DfEE] (1999), Jobs for All: report of the Policy Action Team on Jobs. London: DEE. Department of Social Security [DSS] (1999), Households below average income: A statistical analysis, 1994/5-1997/8. London: The Stationery Office. Ford, J. and Burrows, R. (1999), 'To buy or not to buy? A home of one's own', in R. Jowell et al. (eds.), British Social Attitudes, the 161h survey: Who shares New Labour values?. Aldershot: Ashgate. Ford, J. and Kempson, E. (1997), Bridging the Gap? Safety nets for mortgage borrowers. York: York Publishing Services. Gibb, K. and Munro, M. (1999), Housing Finance in the UK: An introduction (second edition). Basingstoke: Macmillan. Glennerster, H., Hills, J. and Travers, T. (forthcoming), Pulling the purse strings: Formula funding for public services (provisional title). Oxford: Oxford University Press. Hemming, R. and Hills, J. (1983), 'The reform of housing benefits', Fiscal Studies, Vo1.4, No. 1. Hills, J. (1988)' Twenty-first centuv housing subsidies: Durable subsidy and rent-fixing arrangements for social housing, Welfare State Programme Discussion Paper No.33. London: London School of Economics. Hills, J. (199 I), Unravelling Housing Finance: Subsidies, benefits and taxation. Oxford: Clarendon Press. Hills, J. (1998), 'Housing: A decent home within the reach of every family?' in H. Glennerster and J. Hills (eds.), The State of Welfare: The economics of social spending (2ndedition). Oxford: Oxford University Press.
Inquiry into British Housing (1985), Report. London: National Federation of Housing Associations. Kemp, P. (1998), Housing BeneJit: Timefor reform. York: Joseph Rowntree Foundation. Kemp, P. (2000), 'Shopping Incentives ' and Housing BeneJit Reform. London: Chartered Institute of Housing. King, P. (1999), 'Calling time on housing benefit', Housing Today, 7 October. Marsh, A. and McKay, S. (1993), Families, work and beneJits. London: Policy Studies Institute. Oswald, A. (1999), The Housing market and Europe's Unemployment: A non-technical paper (mimeo). Warwick: University of Warwick Department of Economics. Power, A. (1993), Hovels to High-rise: State housing in Europe since 1850. London: Routledge. Power, A. (1997), Estates on the Edge: The social consequences of mass housing in northern Europe. Basingstoke: Macmillan. Wilcox, S. (1998), Housing Finance Review 1998/99. York: Joseph Rowntree Foundation. Wilcox, S. (1999), Housing Finance Review 1999/2000. York: Joseph Rowntree Foundation.
