World Bank FOURTH URBAN RESEARCH SYMPOSIUM Urban Land Use and Land Markets May 14-16, 2007

Transcription

1 World Bank FOURTH URBAN RESEARCH SYMPOSIUM Urban Land Use and Land Markets May 14-16, 2007 Land Use Regulation: Transferring lessons from developed countries R Chiu, B. Turner, C. Whitehead EARLY DRAFT Introduction In the early days of the World Bank s involvement in urban land issues they held a symposium to assist World Bank staff in urban project and design activities setting out the major issues; identifying the analytic approach ; and suggesting priorities and some ways forward. The papers were developed into a text Urban Land Policy: Issues and Opportunities (Dunkerley, 1983) putting the questions into a more transparent framework of economic analysis with the aim of bringing issues into broader academic and policy discussion. The issues raised look frighteningly similar to those that still remain at the forefront of debate: the rapidly rising price of urban land; the misallocation of land resources arising from constraints relating to title, finance and regulation; the misspecification of prices and therefore opportunity costs resulting in incorrect signals to major actors in development; the overwhelming importance of infrastructure provision in both making lad available and in organizing the relationship between urban activities; the lack of potential for capturing land values for development and service provision; and the different mechanisms that could be used in principle and practice to provide the basis for sustained urban growth. The major conclusions of the text were relatively straightforward but fundamental to the ways in which regulation might be used more effectively: urban land systems are inherently inefficient if left entirely to the market - and these market failures are large scale; market failures can in principle be dealt with equally efficiently by regulation, taxation, subsidy or direct intervention which is likely to be the more effective depends on the specifics of the situation and cannot be assessed without examination of the practicalities, and of the ways in which the different interventions interact with one another; interventions have important and often negative impacts on the distribution of welfare but so do market inefficiencies. Trade-offs always have to be made and measures to improve efficiency may often need to be accompanied by measures to improve distribution. In the context of land use regulation per se there was an acceptance that regulation is necessary to address both market failures and inequities. However there was deep concern that existing regulatory frameworks were themselves so inadequate that they were doing more harm than good. In particular:

2 there was little evidence of capacity to simplify existing systems and to relate them more to local circumstances many systems had been transferred from developed countries with little or no changes and so were completely inappropriate in terms of standards, technology, materials etc. As a result the vast majority if building was taking place outside the regulatory framework reducing the potential benefits of planning; systems were inflexible with.no well defined mechanisms for change resulting in out of date regulations which could not readily take account of emerging opportunities; even in countries with apparently strong regulatory frameworks land title and property rights were ill-defined increasing risks and reducing the capacity to invest in the long term physical assets required for effective growth as well as the potential for debt finance; there were high costs of enforcement and inadequate and overstretched administrative systems with little access to funding to develop better arrangements; the potential benefits of regulation in providing a framework for taxation of property and development gains and therefore finance for necessary infrastructure and services was not being realized; systems tended to reinforce inequalities and exclusion by generating large scale benefits to those able to develop within the formal sector while worsening conditions for those without the resources to benefit; it also increased the potential for speculation and manipulation of the regulatory system by those with insider knowledge and adequate resources; and most fundamentally regulatory systems tended to be based almost entirely on constraining development rather than providing a positive environment in which the major actors could respond to opportunity. Changing emphases In practice the starting point for the Bank has tended to be strongly market oriented with much of the emphasis on improving property rights, getting rid of regulatory constraints and encouraging the use of pricing both to improve resource allocation and to generate revenue. In particular much of the analysis has been concentrated on identifying land supply elasticities and much of the policy on attempting to increase these elasticities. An over simple version of their implicit model is set out in figure 1. If left to the market the supply is seen as nearly infinitely elastic. Regulation introduces inelasticites, increasing price, reducing the amount of land made available and reallocating wealth to land owners. The deadweight loss is the triangle abc and is likely to be significant. As such it is a comparative static equilibrium model with no externalities or other market failures. The evidence to support this approach comes from detailed case studies of housing markets across the range of developed and developing countries (refs Malpezzi and Mayo, Malpezzi and Maclennan; Swank et al; Barker etc) These tend to show that

3 supply elasticities vary enormously between countries and, where data are available, between different areas of each country. In particular they show that in Korea, Malaysia, the UK and in most European countries all of which have well developed regulatory environments elasticites are low, ranging from Implicitly in most developing countries formal sector supply elasticities are similarly low although informal sector development may reflect underlying greater elasticity. The exceptions identified are Thailand and the United States with supply elasticities up to 10 in the US. Both of these countries are seen to have strongly market oriented frameworks which favour development. The conclusion therefore appears to follow that regulation is more likely to do more harm than good. Figure 1: Market supply of land price elastic The costs of supply inelasticity are also seen to be likely to increase over time because of the impact of increasing demand figure 2. As demand increases the loss of output and the size of the deadweight loss arising from regulatory constraints also grows (def as compared to abc). Moreover demand has increased not only because of demographics and income growth but also because deregulation of the finance sector has been generally successful allowing enabling demand which had been constrained by lack of finance to enter the market. Finally increasing land prices themselves may further increase demand on the back of expectations of further increases. In this model the costs of regulation can be measured by the gap between the regulated price and the market price and this could be seen as one measure of the negative externalities and other market failures that would have to be identified in order to make the regulation efficient. To the extent that regulation generates befits to land users these would show up as part of the shift in demand and would not be readily separately except perhaps where comparator areas can be examined. These benefits tend to be discounted because there is no simple approach to measurement.

4 Figure 2: Market supply of land price elastic Is this a reasonable model of the impact of regulation? The relevant questions can readily be identified - whether the evidence can be adduced is rather more difficult to answer. First is the market supply actually highly elastic if unregulated? Or do the sources of inelasticity not simply include regulation but also the lack of infrastructure, problems of accessibility, the risks associated with inadequate planning and indeed the impact of agglomeration benefits on different locations. If there is a range of sources which is the most important and does addressing a particular constraint of itself improve the situation? Second, to what extent does the regulated supply curve reflect real negative externalities and other costs of excessive or ill-organised development or is it an inefficient constraint arising from political tensions, slow adjustment or other administrative failures? To the extent it reflects the first is the approach responsive enough to identify the differences in these negativities between particular locations? Can more positive approaches to regulation produce win-win situations where the greater organisation and certainty associated with a well structured regulatory framework supports more efficient production of infrastructure, and makes possible greater investment as a result of the more certain environment as well as possibly enabling the use of other instruments such as congestion charging? Can regulation improve rather than worsen distribution by enabling the provision of public and social goods such as open space and affordable housing as well as effectively supporting the property tax base and the taxation of development gains thus providing an income source for community development and services? While nobody would suggest that regulation can be the panacea for all the ills of urban development it is equally undesirable to reject the benefits of regulation because the

5 extent of bad regulation. Looking at more successful examples in the developed world may help to clarify some of the factors necessary to make land use regulation a more positive force. Lessons from Developed Countries Perhaps the most important issue relates to the elasticity of supply. If the evidence suggests that only the USA among developed countries has been measured as having very elastic supply a reasonable question must be are here reasons why in most countries the supply is quite inelastic? First the supply of raw land in most developed countries is undoubtedly potentially elastic except in countries with extreme constraints. Even Hong Kong and Singapore have not fully developed all the land in their countries. In most countries only a small part of the surface area has been subject to development. However raw land in most instances is not the relevant variable. What matters is developable land with services and accessibility. The most obvious reason why land may not be readily supplied is the provision of infrastructure which normally requires a reasonable clear plan to reduce risk and ensure provision let alone the capacity to finance investment with very long term payouts and often very limited capacity to raise revenue directly from the services such as roads provided. Secondly it is not always possible to replicate accessibility simply by constant levels of investment to that extent the cost of land with similar attributes increases as the urban system becomes larger. At the same time the costs of intensifying the use of the urban area increase with the density of use of land. Both of these factors mean that the cost of expanding equally productive land may rise with scale. The underlying supply elasticity is therefore necessarily more inelastic. In addition as urban systems grow and urban activities expand there is increasing evidence of negative externalities associated both with congestion of transport and utilities and of pollution which impact on health and wellbeing of the population but also on climate and sustainability. These shift the optimal supply curve to the left. Moreover regulation may be the simplest way of taking these factors into account. Three important consequences follow from this discussion: the supply curve of serviced urban land is likely to be relatively inelastic and the more so the more the urban system is developed successfully; as the costs of land are increased by development the price of that land is also higher; regulation may properly reflect the costs of negative externalities and at the same time may increase the benefits achieved by individuals and therefore further increase the price.

6 What is the implication for the costs of regulation? As figure 3 shows once the underlying supply curve is inelastic and the associated price will be less as will be the deadweight loss. This is not to say that regulation could not impose extremely heavy costs if it is too restrictive and does not reflect the social costs of development and if it is not responsive to change. However it is equally possible that effective positive planning can enable the urban system to be better organised and managed, thus increasing the overall benefits to society (figure 4). Figure 3: Market supply of land price inelastic Figure 4: Market supply of land price inelastic

7 What is obvious however is that supply inelasticities, whatever their cause, addressing negative externalities and increasing the benefits of development all increase the price of land (while the extent of land developed is indeterminate). The extent of the pressure on price may also be increased by the existence of agglomeration economies which both increase demand for land and make it more difficult to replicate the benefits of urbanization in systems not generating such high agglomeration benefits. The empirics in this context are very poor. However they are beginning to suggest that agglomeration benefits could be highly corrected with scale diversity and connectivity - all of which put pressure on land prices. Thus expanding in areas where supply is elastic may not alleviate the problems encountered in rapidly growing and successful urban systems. Two other points are worth stressing in this context. First, as incomes rise the valuations of negative externalities in the form of time and wellbeing lost, as well as in increasing problems of sustainability themselves rise further increasing prices if they are properly accounted for. This perhaps suggests that the benefits of well organized planning may be less in developing countries than in the developed economies. However, to the extent that factors of production whether labour or global capital are mobile the extent to which negative externalities are effectively addressed can significantly affect competitiveness. Secondly, the increases in land prices associated with development will impact disproportionately on those groups who do not directly benefit from the increased productivity and incomes notably those providing unskilled and service labour where immigration an hold down wages. Thus there is often further pressure on available housing for poorer groups as well as increasing problems of affordability. Successful planning, just as much as poor planning generate inequalities and the need for additional intervention to provide basic services. It is perhaps for this reason that countries that are seen to have been relatively effective at implementing positive approaches to land use planning such as the Netherlands and France in Europe and Hong Kong and Singapore in Asia have all had a strong commitment to the provision, not only of housing for the poorest but also for affordable housing provision across a wide range of households. Moreover they are all countries where strong government intervention through master planning as well as linking regulation to funding and the development of public private partnership are well developed. Example 1: Hong Kong Hong Kong is a good example of a highly regulated system where the incentives are generally closely aligned with the objectives of increasing development potential. At the same time there is strong governmental recognition of the need fro affordable housing which is enabled by both planning controls and the revenues obtained from the sale of leases. The regulatory process is reinforced by public landownership and by welldefined contractual relationships between the state and developers which provide a range of examples of effective public private partnership working. The planning system involves territorial and sub-regional development strategies and plans determined administratively after public consultation. At the district level almost all

8 developable and developed land is covered by statutory land use zoning plans. Except for a tiny plot of land in the city centre, all land is owned by the government. Land is leased to land users on (generally) renewable long-term tenancies at market prices through auction, supplemented by private treaty grants or tender for special land use purposes. The planning and development of new areas are administered by the government with the participation of relevant organizations (e.g. utilities companies, mass transit companies). All public transport is commercially run. Public utilities companies, except water, are also commercially operated. About 50% of housing is publicly provided directly by government or its agents and is subsidised by government. Two-thirds is rental, with the remaining owner-occupied. Over 70% of private housing is owner-occupied. Within the limits set by the density zonings (there are three density zones in Hong Kong), the traffic capacity, sewerage capacity and other planning controls such as building height restrictions and urban design guidelines, the planning authority usually maximizes the development intensities of new development sites or redeveloped sites, and specifies the intensity in the statutory outline zoning plans. Land premiums charged by the government increase in tandem with development intensities. Between 1989 and 2006, land revenue on average accounted for 10% of the total annual government revenue. Stamp duty levied on property transactions is also increased by development intensities as these usually result in higher buildings, producing more expensive high floor units. The planning process is designed to provide a mechanism for creating better living environments. At the macro scale, the planning of new growth areas (new towns), which translates the territorial spatial strategy and the sub-regional planning concepts into concrete realities, is an active process for creating a living environment meeting contemporary standards, given the constraints and other related policy directives. Planning of this scale incorporates the provision of infrastructure and other urban services especially transport network and community facilities, usually divided into a few phases. At the micro level, the various land use zonings aim towards making a pleasant and compatible local living and working environment. In particular, the zoning of Comprehensive Development Area designated for larger new or redeveloped sites (generally more than 2 hectares) are devised to enable more comprehensive, compatible and complementary provision of residential, retail, community, transport and recreation facilities over neighbourhoods. The land use zoning mechanism enhances efficiency and certainty in the development process. The Notes attached to the statutory zoning plans specify land uses which are always permitted under different zonings, thus providing certainty to land owners and developers and enhancing efficiency in the development process. The Notes also specify land uses which require planning permission which must be granted by the Town planning Board. The outcomes of planning applications are unpredictable and therefore create uncertainty to land owners and developers. Land uses which are not specified in the Notes are not permitted. While this is also a kind of certainty, proposals may be lodged to request changing the statutory land use. The outcomes of such applications are also unpredictable. However, there are statutory time limits for the consideration of planning applications.

9 Undoubtedly planning controls restrict the supply of developable space in Hong Kong (not exactly land supply because of the dominant influence of development intensity in the context of a high-rise built environment) in the sense that planning controls have to combat the desire of developers and property owners to increase land value and profit through raising development intensity above the levels which are socially desirable. Planning does increase development costs and risks where planning applications are necessary, particularly after project commencement. Such costs are often incurred in sites zoned as Comprehensive Development Areas where any change in the approved master plan requires planning permission. Given the scale of development, revisions are often inevitable. Extra costs are also incurred by conditions attached to planning approvals. Planning facilitates the provision of affordable housing in two main ways. The first is the planning and development of new towns to provide serviced sites for the construction of public housing for renting and sale. The second is the extent to which development intensities are increased to enable the supply of public housing. These two approaches are possible because the government owns all land and controls development rights. Moreover the revenues that are generated by the sale of land make large scale provision of affordable housing possible. It thus links the capacity to allocate the land for housing purposes to the resources to ensure affordability. China uses something of the same approach but with clearer trade-offs because profit rates are specified. Land is publicly owned but leased or administratively allocated to land users on permanent or fixed-term leases. Relaxation of development controls on the intensity of use is one of the most important mechanisms to encourage developers to participate in subsidized housing projects targeted at lower-middle income families. While the gross profit rates are fixed in these projects, the development intensities of the commercial portions are increased to allow greater general profit margins. Hong Kong also has much in common with Singapore where government has extremely strong powers to implement plans but equally has a strong commitment to ensuring that adequate housing is provided for all. Land use planning could not be effective on its own in Hong Kong any more than elsewhere. First, planning controls and approval conditions are enforced through the land lease agreements between the government as the landlord and the land user as the lessee, and through building plan approvals executed under the Building Ordinance. Second, state ownership of land reduces the issue of property right in exercising planning controls, particularly in the use of planning tools to enhance the supply of affordable homes. Example 2: the UK Of importance at least in Europe is the general political acceptance of land use planning which makes it easier to implement the controls but more difficult to generate change. The levels of acceptance are extremely high in the UK for instance the legislation of

10 1947 is still consistently upheld by all parties and by the general public even though they recognize that there are costs to themselves. One result of this level of acceptance is that it is possible to use the land use planning system to generate revenues and other benefits for the community from the uplift from enabling development and particular to ensure land is made available for affordable housing. Again the UK provides perhaps the best example because it concentrates on both land and finance (Monk et al; Crook et al) but the general approach is being implemented in many other countries in Europe as well as Australia and Korea and bears some resemblance to local approaches in the USA. In the UK throughout the 1980s there was growing use of planning powers to require developers to provide not only the investment necessary to offset any immediate externalities associated with the site but also to provide for the broader needs of the community arising from development often in the form of a levy for transport and education. In 1990 the Town and Country Planning Act was enacted which included section 106 (s106) which enabled broader negotiation notably with respect to affordable housing. As long as a local authority has in place a local housing needs assessment which shows the need for additional affordable housing they can require residential developers to make a contribution towards that provision. This contribution will normally be required on the site, so supporting the mixed income communities agenda. It can be in the form of social rented housing and or low cost shared ownership accommodation. Depending on the extent of planning gain and negotiations the developer may be required to provide the land or indeed the housing free of charge or there may also be government subsidy available to make it affordable. In the more pressured parts of the country the contribution will normally be of the order of 20 25% of output; in London the aim is between 35 and 50%. Overall the requirement enables significant developer contribution which itself reduces the sale price of land so that at least in principle the tax is paid by the land owner. The process is only possible because of the nature of the land use planning regulations and because of the planning gain arising from permission but as argued above this gain is almost inevitable in successful urban systems. The approach can also be used effectively to increase densities of development and therefore to support the sustainability agenda. In the UK there is a requirement that 60% plus of all new housing be located on brownfield land much of which will be being redeveloped at higher densities. Technically the use of s106 is a separate instrument but in practice because of the on-site requirement the mix of privately provided housing also changes with a resultant increase in density. In Australia - and China density is used as an element in the negotiation to enable larger contributions to affordable housing to be achieved. Three Caveats No one could suggest that either of the examples above are first best solutions indeed the World Bank itself has been known to define the UK system as Stalinist and there is currently much debate about how to increase supply elasticities without losing the benefits of planning. Rather they are approaches to bringing together different

11 instruments in such a way that development compatible with successful urban growth and limiting negative externalities can be achieved. They are also means by which funding can be provided to contribute towards the provision of infrastructure and particularly of affordable housing. They both depend on political commitment and on government involvement. But they suggest ways forward in circumstances where the starting point is an existing planning system rather than an unregulated. Experience of past attempts at reform in these circumstances shows that simply removing particular constraints has little impact on behaviour and outcomes so it is important to build on positive partnership approaches. Evidence also suggests that as incomes rise the benefits of moving towards a better organized regulatory system are likely to increase. Second, experience suggests that in most regulated systems the most effective means of achieving a large scale step change in investment in housing requires significant government intervention either to provide publicly owned land or to reduce the price of land for development and to provide services. The skill is then in obtaining a fair return on that investment through increased revenues into the longer term. Effective planning is a necessary condition for this to be possible. Third, the most important emerging problem is that of the effective organization of regeneration and redevelopment especially in environments where owner-occupation and therefore fragmentation of land ownership is prevalent. To generate higher valued higher density developments in these circumstances requires the effective use of planning powers as well as appropriate compensation mechanisms. It is undoubtedly the major challenge to the land use planning system in the twenty first century. Conclusions First bad regulation i.e. regulation that is not related to the social costs and benefits of development, which cannot adjust to changing circumstances and appropriate technologies, and which cannot enable development to take place remains bad. Similarly increasing the elasticity of supply ceteris paribus is desirable. In many circumstances in the developed economies the incentive structures to enable development are lacking as those who already have a stake in the system lose out as a result of providing additional land and housing. Equally however market systems appear to become less flexible as urban systems develop. In this context the issues of NIMBYISM no longer seem to be so strongly associated with traditionally planned economies but apply as much to more market based economies such as the USA at least in the more desirable areas. It is also well worth remembering that regulation is the easiest form of government intervention to introduce and is often done as a knee-jerk response to crisis. But modifying existing regulation is often far more difficult if only because those who gained from its existence are usually more powerful. Second the benefits of well organized urban systems which provide a good living environment as well as increasing productivity in crease with incomes. At the same time the costs of the negative impact of growth such as traffic congestion, over-utilisation of

12 public services and pollution also increase. Therefore the benefits of getting the regulatory framework to operate effectively increase with success. Third, land use planning can provide a strong base from which to tax development gains and thus support infrastructure provision either directly through the revenue raised or by reducing the cost of land through the imposition of planning requirements. Finally land use planning cannot operate effectively except within a suite of complementary policies, which must include instruments to offset the negative distributional impacts of even the most successful planning regimes.