LAND VALUES AND THE IMPLICATIONS FOR PLANNING POLICY
CHAPTER TWO: REVIEW OF RESEARCH EVIDENCEIntroduction
2.1 While the main focus of this review of research evidence will be the impact of land values on urban form and the consequent implications for planning policy, planning policy is itself one of the determinants of land values. In this review of research evidence, therefore, we focus on both:
- the relationship between the price of development land and urban form; and
- the impact of planning on land values.
URBAN ECONOMIC THEORY- PERSPECTIVES ON LAND VALUES AND URBAN FORM
2.2 We open this review with an overview of relevant work from the discipline of urban economics on the processes which impact on land values and urban form. The academic literature on urban economics describes a number of processes in urban land markets which may help to explain why, for example:
- land values vary spatially and over time;
- why densities vary in different parts of conurbations;
- why particular uses develop in particular parts of cities;
- why particular income groups may choose to live close to or distant from their place of work.
2.3 The early theoretical literature 1 describes the forces and processes which may lead to there being not just a rent gradient from the centre of a conurbation, but also a density gradient, and the reasons for the two being inter-related.
2.4 The intellectual basis of the standard urban-rent model with its concentric circles, which forms the basis of much theoretical work in urban economics, stems from the agricultural location theory of the early 19 th century writer von Thünen, whose work provided a hypothesis which explained inter alia why the most intensive crops were produced closest to the market. Modern urban land use theory, which forms the core of urban economics, is essentially a revival of von Thünen's theory of agricultural land use. 2
2.5 Subsequent writers adapted von Thünen's theory to an urban setting, emphasising the role of competitive bidding for land in determining urban land uses and the influence of accessibility on land values. In a city with a single centre, space will be used most intensively in the core and the density of use will tend to decline with increasing distance from it. These gradients may also be observed in population densities, urban land values and employment per acre of land.
2.6 Writers such as Park and Burgess, influenced by ecology, suggested that human beings compete for scarce resources such as land and raw materials, with the aim of establishing spatially disparate urban environments to satisfy their different economic and social needs. As in ecology, the boundaries between such environments are constantly changing. 3
2.7 Alonso's 4 development of the bid-rent function formalises the trade-off between accessibility and land costs. Each activity or land use has a family of bid rent curves which shows what a given activity is prepared to pay at each site. The activities with the steeper bid rent curves capture the central locations because they are prepared to pay more for central sites. Households also have a bid rent function - a trade-off curve between housing costs and journey to work costs, which generally assumes that the household has a fixed budget that it can allocate to some combination of these two items. In fact however, the availability of mortgage finance may constrain that set of preferences and households may only be able to choose up to a certain maximum housing cost point.
2.8 Alonso's bid-rent concept implies that with an increase in urban population and/or an increase in total urban income, the demand for land would increase - raising bid-rents throughout the urban area. This in turn, would result in each land use invading the next outer zone. 5
2.9 Different income groups will have different bid rent functions - it is commonly assumed for example that the poor have steeper bid-rent functions than the rich, because it is necessary for the poor to live near work at high densities to economise on transport costs, whereas the rich have a high income elasticity of demand for space.
2.10 The pattern in many Western cities where the very rich choose to live either close to the centre or well out in the suburbs can be explained by the rich having a convex bid-rent curve (Fig 2.1) - ie. they either want to live very close to the city centre for access to its amenities or in order to minimise the time costs of reaching downtown offices or in the outer suburbs, where they may enjoy the advantages of more space, access to open country, or a quasi-rural lifestyle.
2.11 There is some empirical evidence that increasingly strong demand for city centre living is leading to a doughnut effect 6 in some British cities - most notably in Birmingham and London, where the Town and Country Planning Association has argued that the suburbs have been left behind in the rush to improve inner city areas. Areas such as Selly Oak, Northfield and Erdington in Birmingham and previously prosperous areas such as Surbiton, Barnet and Colllier's Wood in London are suffering from failing infrastructure and an exodus of retailers to the city itself or to out of town locations.
2.12 If these academic models are dynamised they suggest that:
'growth will lead to a pushing out and widening of concentric zones of activity, and land uses dominant in one ring by a process which the urban ecologists have called 'invasion - succession' …………….some spatial characteristics of urban expansion are obvious and fairly universal. Urban growth will normally be associated with both upward and outward expansion. Near the city centre, low structures will tend to be replaced by higher ones and there will be attempts to encroach on open spaces and build on any spare plots. Central densities (with the probable exception of residential density) will rise, and the central land will be used more intensively. At the same time, however, the settled urban area will expand outwards. One of the main reasons for this is that efficient production of goods requires new land as an input. Another is that the provision of new housing for an expanding city population will require (and be cheaper on) new land, especially as, as is often the case, rising income levels are associated with a demand for more living space7.'
2.13 Empirical work in the US led to a critique of the concentric zone model, which was found to offer a poor description of the structure of many cities. Hoyt's radial-sector theory, developed in the inter-war and immediate post-war period, describes how in many cities one or more high rent sectors shaped like pie slice wedges tend to develop along radial lines from the centre to the periphery. High rent sectors tended to develop in areas which offered particular attractions - primarily environmental.
2.14 While these theoretical models all offer useful insights into certain of the processes which determine urban form, they are challenged by the complexity of real cities, where so many factors - not solely economic - impact on land use decisions and the development process. The fact that cities are increasingly polycentric, with many subsidiary centres of employment, makes the models hard to apply to the real world.
2.15 In the UK, the impact of the planning system is of course a powerful determinant of both land uses and densities, which does not always take into account the natural tendencies of the market - the forces which underpin the theoretical models. For instance, the City of Edinburgh's recent masterplan for Craigmillar produced by Piers Gough proposed high density housing on a low value, relatively peripheral site, motivated by architectural considerations and arguments that higher density provision would result in a somewhat higher density of public transport provision. This was contrary to the preferences of the housebuilders, who saw this as a location where lower land prices would make it possible to provide lower density housing, which would be difficult to provide elsewhere in the city - housing of a type which housebuilders can now generally provide only on lower value sites well outside the city boundary.
2.16 Modern urban economics uses more complex models of locational choice, which reflect the fact that the decisions of households in particular are not a simple trade-off between space and accessibility - the trade-off is in fact a three way trade-off with three basic factors being considered: accessibility, space and environmental amenities.
"Accessibility includes both pecuniary and time costs associated with getting to work, visiting relatives and friends and shopping and other activities. The space factor consists of the need for some land as well as the size and quality of the house itself. Finally, environmental amenities include natural features such as hills and scenic views as well as neighbourhood characteristics ranging from the quality of schools and safety to racial composition."8
2.17 More recent work in urban economics seeks to model the impact of these other factors - the neighbourhood externalities which explain why different parts of the city which are equally accessible from the city centre may have such different land prices.
2.18 For example, the work of Cheshire and Sheppard 9 involves the creation of models which attempt to deal with the considerable complexity of the determinants of urban land prices. Hedonic price models are used to explain intra - urban variation in the price of land. One of the factors which impacts on land prices is the planning system, and some of Cheshire and Sheppard's work attempts to quantify that effect 10. This aspect of their work is discussed in more detail in later parts of this section.
2.19 Cheshire and Sheppard's work recognises the many factors which impact on the price of an individual plot of land. They state that :
'It has long been recognised that housing (and for that matter, commercial or industrial buildings) is a composite good. The price that is paid for a house reflects the various characteristics of the house - its floor area, for example, or the facilities it enjoys, its age and design. A house, however, is not only composed of characteristics relating to its structure but also of characteristics determined by its location. The latter include the classic element of urban economic models, accessibility to the employment centre. There is another set of location determined characteristics, however, including the quality of local public goods and of the microenvironment, the characteristics of the immediate neighbourhood, or the amenities (and disamenities) which the location provides access to.
Since the set of location specific characteristics is tied to the parcel of land on which a house stands (rather than to the structure itself) what this implies is that land also is a composite good. Theory explains the price of land as 'pure space with accessibility' but the price of any parcel of land as directly observed incorporates the values of all the location specific characteristics with which it is endowed. Plots of urban land turn out on close examination to be seething with these location specific and valuable characteristics which are in sum so valuable that typically the market price of any plot is dominated by them rather than by its value as space with accessibility.'
THE IMPACT OF PLANNING ON LAND VALUES
2.20 To review the assertion that planning policies are impacting negatively on urban environments through their impact on land values, we review the academic literature on this topic. We then go on to review the extent to which the literature suggests that the planning system in Scotland has focused development in inner city areas and whether this is responsible for driving up land values in these areas. Planning may also be expected to have positive impacts on land values where it restricts demand in certain locations in order to focus demand on areas where demand is weak.
Planning Policy Trends
2.21 Until the 1990s, our view of the long-term impact of planning since the post-war years would have been that it had tended indirectly to support trends of decentralisation. Dissatisfaction with urban environments had encouraged many to move to suburban areas where much of post-war private house building has been concentrated. Such housing development was facilitated by planning policies. Indeed, in the immediate post-war years, public policy initiatives led to the development of the new towns and peripheral estates, and thus actively supported the decentralisation of both housing and, in the case of the new towns, employment. Generally, the planning system tended to accommodate the shift outward, as statutory plans tended to make provision for land allocations in areas where population growth was projected 11 - on a 'predict and provide ' basis. In more recent years, however, the planning system has acted as a constraint, by responding to changing demand with a lag, so that land supply has often failed to match current demands.
2.22 More recently, government policy has shifted to halt the depopulation of urban centres whilst encouraging sustainability. The report of the Urban Task Force contains one of the most succinct summaries of that policy shift 12 in England. (In Scotland, the equivalent review of urban policy, the Cities Review, will be published next month.) Policies have attempted to focus development in cities. The resulting change in emphasis in planning policy combined with the continuing limitations on developing green belt land has led to an increasing proportion of development being concentrated on brownfield sites. There is also evidence, cited in a recent Scottish Executive survey, that where pressure has become too great, development has moved out beyond cities and their surroundings 13. In the case of Edinburgh, new housebuilding has moved out beyond the green belt to West, Mid and East Lothian and to Fife.
2.23 Although the success of this policy has been noted - in the Glasgow and Clyde Valley Structure Plan area 60% of new development has been on brownfield sites - there has also been much discussion of the continuing plight of some inner city areas which have failed to attract development interest, despite the best efforts of planners. It is acknowledged that the statutory planning system can act as a negative force, able to prevent development effectively, but often powerless to stimulate development in the face of unfavourable demand conditions.
2.24 Demand, whether originating from the market or whether initially stimulated by government subsidies, is imperative for development in city centres. Where favourable market conditions for development have existed, planning has succeeded in channelling growth into built up areas yet many inner city areas still grapple with the problems of stimulating demand even when planning policy is supportive of development 14.
Planning and Land Values
2.25 There is a body of literature relating to the impact of planning on land values. Planning, by limiting supply, is generally acknowledged as impacting on land values, although the extent of this influence is debated. The arguments were summarised in a recent report for the Scottish Executive: 'The Role of Planning System in the Provision of Housing' 15.
2.26 Those who emphasise the role of supply factors over those of demand, such as Evans 16, argue that planning regulations can push up the price of land if they constrain the supply of land and thus impact on the elasticity of the housing supply. In these circumstances increase in demand results not in an increase of housing but in an increase in house prices. This is compounded by the spatial segmentation of housing markets. The planning system regulates the supply of land thus preventing house builders from responding to demand. Over time, inflated house prices and high land values will tend to result in a shift in supply towards properties which consume less land.
2.27 In contrast it has been argued by those who emphasise demand side factors such as Grigson 17 that the planning system cannot push up prices of land or housing other than in certain market conditions. House prices are largely determined by demand because the sale of new houses makes up such a small part of the market. Furthermore, the supply of housing is relatively fixed in the short run due to the time taken to complete new developments. However, at the margin, new developments do impact on price.
2.28 Empirical studies have tended to support the supply side argument and suggest that planning processes do have an impact on land values and house prices. Comparative studies of areas of strong and weak planning control, such as those by Cheshire and Sheppard 18 have tended to conclude that such differences in policies have resulted in price differentials. A study by FPD Savills on land values in London, for example, concludes that land for office development may be cheaper due to more flexible planning constraints. 19 In terms of recent Scottish experience, it is clear that limitations in the supply of residential development land in particular have resulted in rising land values in Edinburgh and the Lothians 20.
2.29 The only major piece of empirical work on land prices in Scotland 21, a study carried out by Pieda in 1986 for the Scottish Development Department, showed that both demand and supply factors were significant. The statistical analysis of variation in land prices provided some evidence that constraints on land supply resulted in increased house prices, but demand factors, represented by the level of earnings, unemployment and owner occupation, were also important. In each of the equations tested, the level of earnings proved to be the most significant variable, followed by the measures of land supply, such as the amount of land with detailed planning permission per capita.
2.30 The 1986 Pieda study showed that central Edinburgh land prices - at that time also the highest in Scotland - were much higher than in the district council areas surrounding the city. In Glasgow, by contrast, certain of the adjoining district council areas (high status suburbs such as Bearsden and Milngavie and Eastwood) had land prices which were almost as high as those in the city itself. The report's authors speculated that this sharp fall off in house prices as one moved out from Edinburgh could be a consequence of the 'encircling' nature of the green belt in Edinburgh. The alternative hypothesis - which is that land in the surrounding districts is not regarded by builders as a substitute for city sites, reflecting house purchaser attitudes - is also put forward.
2.31 Of particular relevance to this study is the apparent impact of recent changes in the planning system on the supply of development land and resulting land values. The Property Market Report based on analysis of the recent UK property market by the Valuation Office, argues that:
'PPG3's emphasis on the importance of social housing provision and the requirement that 60% of new housing should be provided on brownfield land has concentrated developers minds and any land with existing planning consents is much sought after. This trend has caused significant demand for land in many areas with a consequent upward pressure on values.'22
2.32 Another potential impact of the planning system on restricting supply may be through the operation of affordable housing policies. This might happen in one of two ways:
- In London, it has been argued by the RICS 23 that affordable housing policies are so demanding (because the policies suggest that a very high proportion of housing is to be affordable) that developers may simply shelve schemes.
- In Edinburgh, developers have argued that because the supply of housing sites is so limited, the Edinburgh policy, which focuses particularly on social rented housing, will have the effect of limiting the amount of land which is available for housing for sale. (This argument however is not about housing land overall, only about the amount which is available for housing for sale.)
2.33 In the Property Market Report the role of demand is referred to and although most studies acknowledge the role of planning in influencing supply and thus the price of land, they also point to the significance of demand in determining final values. After modelling the housing market, for example, the Scottish Executive report concluded that even if a large amount of land was to be released via the planning system a correspondingly large increase in housing supply would not occur. Furthermore, such a large release would have a small impact on house prices 24.
2.34 Urban economic theory tends to support our own conclusion about the causes of the sharp increases in land prices, as for example in Edinburgh, which is that both supply and demand factors are important. Theory suggests that a growing city population would, even in the absence of a planning system, result in increased demand for residential land which in turn will increase the price of land everywhere, increase densities and push the urban fringe outward. The more the planning system restricts the outward shift of the urban fringe, or any alternative source of land supply, the more intense is the rise in land prices likely to be. Since we have no evidence to support the view that the planning system has increased its constraints on land supply in Edinburgh, it is likely that the major part of the steeply rising land prices in Edinburgh may be accounted for by increased demand.
2.35 In areas of low demand where efforts have focussed on increasing market interest over a long time period, positive land values have finally been achieved. Recent evaluation work carried out by DTZ Pieda has also shown that urban policies may be effective in altering patterns of demands and thus local land values. However policy-makers need to be aware of the length of their time commitment to that policy - and of the resources needed to support it 25.
2.36 Variations in the nature of demand also influence resulting patterns of land use. Consideration of development processes leads us to conclude that it is a strong demand for a restricted supply of sites which generates high land values and leads developers to plan higher density developments in order to win sites.
2.37 When high demand combines with restricted supply and a planning system that actively encourages intensive housing development, high land values for high density sites can result. The Property Market Report concludes:
'The message in PPG3 to avoid the inefficient use of land and provide more intensive housing development in and around existing centres and close to public transport nodes has led to greater density of development in planning consents particularly on sites for the construction of flats, and in consequence there have been some exceptionally high values paid for such sites.'26
2.38 Land values are clearly the result of complex, localised markets. Moreover, it is worth noting the role land values have had in influencing planning. Monk et al, for example, use empirical evidence to argue that although the planning system constrains the supply of land in general, it does respond to demand pressures subject to a time lag 27. In terms of Scottish experience, DTZ Pieda has argued previously that historically the planning system has tended to accommodate and adjust to market pressures while tending to protect areas with restrictive designations (AGLVs, green belt) from development 28. The adjustment to the market has been in largely plan led areas where projected substantial population growth has resulted in higher allocations of development land.
LAND VALUES AND URBAN STRUCTURES: EMPIRICAL EVIDENCE
2.39 In assessing the significance of land values for urban structures, we have looked for evidence in the literature on the impact of land values on:
- the form of development, urban densities, the structure of cities and towns;
- the sustainable development of cities, including the issue of whether 'balanced communities' are being put at risk;
- environmental quality.
2.40 The factors which combine to produce high land values tend to influence urban structures. Both theoretical work and empirical studies suggest that density increases as land values rise. Higher prices for development land encourage infilling of vacant plots and increased densities. Higher house prices mean that these properties are accepted by many consumers 29, who must trade-off accessibility for space, recognising that they can only have the space they might prefer at the cost of longer commuting times.
2.41 DTZ Pieda has argued that because land values are effectively a residual, greater demand for land often results in higher densities in general. In a situation where there is competition for sites and strong housing demand, developers may try to maximise the number of units on a site, so that they can maximise the residual value, and thus the bid price of land.
2.42 With the increased demand for city centre housing from city centre workers in certain urban locations, the competition between different land uses may be quite finely balanced. The study by Savills shows how the changing hierarchy of property values through different property cycles may make mixed-use development commercially advantageous. For tertiary out of date or off-centre office buildings in a range of locations in London, Savills looked at three different end uses - office, residential and hotel use. The margins between the use which generated the highest Gross Development Value and land value ( office, residential or hotel use) were quite small and may shift quite rapidly over time. On an identical site, the economics of development will favour different uses at different times and land values will change according to property values, developer behaviour and other factors. The study drew attention to the way in which, for example, affordable housing policies can easily tip the viability of a site in favour of offices if no such provision is required for commercial uses.
2.43 Savills argue that in the face of this type of uncertainty, development risk may be ameliorated by building more than one use on a site, which in turn may reduce the risk related margins required by bankers. There has of course been significant growth in mixed-use developments in the UK, attributable to a whole range of other factors - the market has discovered that mixed-use schemes may be more varied, interesting and vibrant than single use schemes, and it has been demonstrated that a mixture of land uses can create a critical mass over and above the sum of the parts. The diversification of risk is an additional factor which may lead to a mixed-use development being favoured. 30
2.44 It is likely that in some urban areas, high land values have been an important factor in encouraging the movement of manufacturing industry out of city centres and the inner city to surrounding districts. Indeed, in Edinburgh not only have industrial uses been displaced by housing, but there are even some sites where housing use has succeeded office use in fringe urban locations.
2.45 For some industrial or service users, the opportunity to release the capital value of the asset they occupy and re-invest the proceeds - or retire to sunnier climes! - may be more of a factor in the decision to sell the site than the fact that they have outgrown their present site or wish to relocate to a more convenient location.
2.46 Although trends in the relocation of manufacturing industry have been documented - by Turok and Edge in the case of Glasgow for example - it is relatively unusual for it to be high land values which are the main drivers in such relocation 31. Other highly important factors include the changing needs of manufacturing production as well as policy initiatives to attract manufacturing into areas outside conurbations and to capture footloose industry by offering greenfield sites with excellent road transport connections. GVA Grimley and Business Strategies have argued in a recent report that ' sites on the edge of urban areas or around 'clusters' of manufacturing estates are preferred, rather than inner urban areas surrounded by residential uses with poor access arrangements' 32. However, in Edinburgh and in West Lothian, there is a considerable amount of recent anecdotal evidence to suggest the proprietors of business land are actively seeking to achieve residential values on their sites.
2.47 The role of land values in determining patterns of residential housing is in some respects counter-intuitive - it might be thought that low land values would have the potential to stimulate residential development. However, the reverse is the case - low land prices in towns and cities tend to reflect fundamental weaknesses in the market for housing in these cities, and a profound reluctance of housebuilders to take up sites, at least in certain parts of these cities. It is by no means the case that housebuilders will perceive low land prices as a positive factor, enabling them to build more cheaply - rather low land prices reflect the builders' perceptions of the strength of demand for housing in these areas.
2.48 Scottish experience tends to suggest that employment and population loss to surrounding areas is more commonly associated with low, rather than high land values. The areas which saw the biggest percentage decline in population between 1991 and 1996 were Glasgow (-2.4%), Dundee (-3.8%) and Inverclyde (-4.9%) yet this has not been land or house price driven. All these places have relatively low land prices. This has been a long-term trend which has occurred in a variety of cities, suggesting that rather than high land values, a complex pattern of consumer preferences - particularly relating to the quality of residential environments and local amenities such as schools - and employment opportunities has been at work 33.
Sustainability: The Case for Higher Urban Densities
2.49 The increase in urban densities which may be associated with high land values is viewed differently from different perspectives. Those who support compact urban forms see increased densities leading to a range of positive results. There is a powerful architectural/urban design lobby, for example, that associates higher densities with improved urban design 34. This view was put forward in the report of the Urban Task Force. High density neighbourhoods are said to contribute to improved energy conservation, transport and service provision - key aspects of the sustainability campaign. This argument is illustrated in a highly relevant paper by Professor Sir Peter Hall which inter-alia, summarised research by the Bartlett School of Planning which suggested that higher densities result in better public transport; support of a wider range of services; decline in journey distance and more localised contacts and activities which reduce the need for travel 35.
2.50 However, this view is not universally held. Those who attack what they term 'town cramming' argue that higher density residential patterns effectively act against consumer preference to force people to live in housing which they do not want to live in. Alan Evans argues that consumers are steered towards smaller properties and higher density living 36. Indeed the studies quoted in the summary of the debate by Peter Hall suggest that the British tend to aspire to live near the countryside, put up with the suburbs as an alternative and are averse, in general, to urban living 37.
2.51 It is worth noting that recent analysis of the UK property market suggests that demand for individual dwellings remains 'insatiable and ubiquitous…in both the more prosperous and the more deprived regions'. 38 Even in Scotland, where urban form has been historically more continental and thus higher density, the English garden city ideal has had its impact.
2.52 It may be wrong too to conclude that the one person households who make up a very high proportion of the total growth in household numbers will want to live in small flats. Research by the Housing Research Foundation for the National House Building Council 39 suggested that in the fastest growing group of households, one person households in the 35-54 age group, most will be house purchasers, who will have clear preferences about the kind of residential environment they want to live in. Most want two bedrooms, and some want three or more. Only 10% would prefer a flat to a house and fully one third would not even consider a flat. This group is looking for parking for their cars and private open space - not a large garden, but a small garden or patio.
2.53 Of most interest in terms of densities and land values are the speculations of Hall on the optimum density for residential neighbourhoods in cities. He argues for a minimum density of 37 units per hectare but acknowledges that in areas of high demand such as Islington and Chelsea in London or Pacific Heights and Russian Hill in San Francisco a much higher density - anything up to 120 units per hectare can be sustained without sacrificing quality of life 40.
2.54 Similarly, Lord Rogers has pointed out that Barcelona has an average density of about 400 dwellings per hectare, and areas such as Bath, Edinburgh's New Town and Bloombury and Islington in London have an overall density of 100-200 dwellings per hectare. 41
2.55 Nicholas Schoon, the former environment correspondent for The Independent, was commissioned by the Joseph Rowntree Foundation to write about the obstacles to the urban renaissance and how they might be overcome. He reviews in some depth the need to make higher densities and mixed-use neighbourhoods fit with people's aspirations for good-sized homes and gardens. 42 He proposes a high density, low rise block predominantly of houses with gardens which achieves a density of 60 houses to the hectare. At this density, he argues, basic local services within a few minutes walk and good public transport can supported.
Urban Densities and Travel to Work
2.56 The implications for transport patterns and the environmental impact of compact urban forms are by no means clear. We would caution against an over simplistic view of the causal relationship between development patterns and unsustainable commuting. The literature on job search and commuting makes it clear that this is a complex area, where many other factors influence the decision to commute and whether to do so by car or by public transport.
2.57 Farrington found that car ownership in Scotland was increasing faster than elsewhere, and that car and rail use has increased, with bus use falling. Journey to work is strongly correlated with socio-economic groups. Those in professional and managerial occupations are more likely to commute to work, and the composition of the labour force has been changing so that an increasingly high proportion of workers are in professional and managerial categories 43.
2.58 A report by BSL in 1998 44 showed that the proportion of the workforce who are commuters between local authority districts rose steadily in the early 90s in Central Scotland. The proportion of employees who were in-commuters rose from 28% to 33% in Edinburgh and from 40 to 44% in Glasgow over the 1991-1996 period. The rise in out-commuting was also substantial in Glasgow.
2.59 Breheny et al argue that their research into these issues ' casts doubt on the orthodoxy that increasing building densities will necessarily reduce travel in towns and cities' 45. Breheny 46 argues that while enormous momentum has built up behind the idea of promoting more housing in urban areas, the question of whether the changing geography of jobs will aid or hinder this approach has been neglected entirely - even in the Urban Task Force Report.
2.60 Breheney's report focuses on the changing geography of employment, and argues that throughout the UK, there has been substantial decentralisation of employment, driven by a number of factors, with large conurbations the major losers of employment. However, Breheny rejects the view that it is this urban decentralisation which has fuelled longer work journeys - he believes that they are more likely to be the result of increased mobility.
2.61 A theme throughout the report is the growing trend for workers and their families to put down long-term roots in a location for lifestyle reasons and assume that they will cope with job changes by commuting. Breheny calls this 'the roots effect' - or a variant - the 'professional roots effect', where dual career professional households choose an optimal location from which to undertake two sets of commuting trips. The location will be chosen for its attributes as a desirable residential area - attributes including environmental quality, freedom from crime and good public services.
2.62 The two questions which Breheny seeks to address are:
- "to what extent have home location and job location become uncoupled?" and
- "to what extent is the geography of jobs susceptible to intervention?
2.63 Breheny believes that the answer to the first question is - to an increasing degree, as a result of increasing real incomes, car ownership, the increased specialisation of the workforce and the higher proportion of professional workers.
2.64 In relation to the second question, he is sceptical about the scope for creating new manufacturing employment in cities, and suggests that although it may be desirable to raise employment within conurbations, there is a risk that the creation of employment next to disadvantaged areas will not lead to a reduction in unemployment in those areas because the jobs are taken by others from elsewhere in the conurbation. (This has recently been reported as being the experience at the Clydebank Business Park (formerly the Enterprise Zone)). Breheny argues for an intensification of training programmes to allow the most disadvantaged of the unemployed to compete.
2.65 Further recent research on moving for job reasons has confirmed the role of dual career partnerships in constraining job-related moves. 47 While this work focuses particularly on migration, it is also highly relevant to the propensity of households to relocate to achieve proximity to the new place of work. Both US evidence and this study confirm that having an employed wife is associated with reduced male migration - as the Warwick team put it, 'men are more likely to be 'tied stayers' if their wife is working.'
2.66 Those who argue against blanket measures to encourage compact urban living maintain that increased densities fail to save land anyway. Peter Hall states that changing household structures mean that ' dwelling densities will not yield person densities that we could achieve 40 or 50 years ago' 48.
2.67 There is limited consideration of the impact of the increasing number of mixed land use developments on sustainability issues. Will these, for example, discourage the use of cars or will home and work remain separate for the majority and car use patterns remain unchanged? There is also little consideration of consumer preferences in the future. Although low density living in a suburban, low density or rural environment remains attractive for many, there are some social groups who may welcome higher densities. Single person households are predicted to become the driver of growth in household numbers in future years. The preferences of this group may also affect the implications of higher densities and the definition of sustainability.
2.68 High land values may suggest that there is a risk that balanced communities may not be attained in new developments. If land values are partly determined by the value of properties ultimately built on the site, then high land values imply the production either of very dense developments or of expensive properties only accessible to certain groups of the population. This implies the creation of communities where there is no housing affordable by lower and middle income groups. The impact of higher housing prices which may threaten inner city services indirectly through the exclusion of public sector workers such as nurses and fire-fighters is well documented 49.
2.69 In the past, many public sector workers relied on the safety net of local authority rented housing, which pooled rents on new properties and older properties whose debts had virtually been written off. In this way, the local authority sector could afford to compete on the open market for land and to build new houses. The advent of Right to Buy has undermined the foundations on which this system was based and as a result, social housing can now only be provided through subsidy (HAG or similar) or through subsidies created by planning gain.
2.70 Yet the localised nature of areas in high demand - these areas may sit next to areas with very different market characteristics - raises questions about the prospect of increasingly segregated communities. There is little research on the household type, economic and demographic profile of urban populations in newly developed areas. These features are important in determining the degree of 'balance' of the community.
2.71 The unfettered workings of the market appear unlikely to lead to greater balance and integration between social groups - indeed, it is only through such policies as the requirement to provide affordable housing as an integral part of new housing for sale projects that lower income groups will have any representation in the most affluent areas of our cities.
Employment Decentralisation and Social Exclusion
2.72 Land values, in so much as they impact on the movement of manufacturing industry out of cities, may be argued to impact on the sustainability of urban environments. It has been argued by some that the decline of manufacturing employment in inner city areas is a major cause of the substantial pockets of unemployment which are a feature of many cities - although it is by no means clear that rising land values are the main reason for declining manufacturing employment in the major industrial cities.
2.73 In the older industrial cities where population has been tending to decline, the fall in employment has exceeded the contraction in population. The result has been high unemployment in older urban areas such as Glasgow, Liverpool and Manchester and a concentration of unemployment in their public sector housing estates. In some cases this has been accompanied by the emergence of low demand for housing.
2.74 Although there may be rising service employment in city centres and in new commercial developments, many are excluded from these new employment opportunities - for example male blue-collar workers. Turok and Edge call for a determined effort to resurrect inner city manufacturing employment in response.
2.75 However not all agree that the loss of manufacturing employment from city centres, has been responsible for pockets of high unemployment within cities. Morrison an American academic geographer argues that urban labour markets are not 'spatially segmented' but 'seamless'. Commuting within conurbations means that within broad skill groups employment opportunities tend to be equalised across the city as a whole. Since above all else unemployment rates vary by skill level, residential areas with relatively more unskilled labour will manifest higher unemployment rates. Demand for unskilled labour has decreased in general so that pockets of unemployment are not a response to the decline in employment in a particular locality but the result of the workings of the housing system, which tends to concentrate those with the least ability to compete in the labour market in particular areas 50.
2.76 In any event, the overall decline in manufacturing employment, which has been a long-term trend throughout the UK for the last decade or more makes the policy prescription of Turok and Edge hard to follow. In our view, it is the overall decline in employment opportunities for those with low skill levels and low educational attainment, combined with the way the labour and housing markets interact, which are the main factors behind the pockets of unemployment found in certain low demand housing areas.
2.77 It is possible that some groups will be stranded in low demand but affordable housing in inner city areas whilst commuting out to work located on the fringes of conurbations. There is little research on this subject although there is evidence of an increase in out-commuting from areas such as Glasgow 51. This has different implications for public transport needs and car usage than those that arise from commuting into urban centres.
Service Provision for Balanced Communities
2.78 Where residential land values are very high, it may seem likely that local services of various kinds will be 'bid out' of the land market. Concern here is that housing areas may become more homogeneous and less well provided with services. Yet there is no clear research evidence on this point. It is observably often the case that areas with the poorest range and quality of local services are not those of high land value - but low land value neighbourhoods, where retailers perceive market opportunities to be limited. 52
2.79 This is an area where the planning system may have an important role to play, because without commercial zonings, high land value residential areas might develop with inadequate land use diversity and thus inadequate local services. However, it has been argued that current trends in the character of local plans make it unlikely that they will be sufficiently fine grained to reserve sites for community uses such as GPs, churches, youth clubs and scout huts. Current pressures are to ensure that local plans are more up to date than they have been in the past, and this makes it less likely that they will include such detailed allocations. This is a debatable point however - it is not particularly hard to ensure that zonings cover existing service uses - the difficulties are rather greater when it comes to ensuring that space is built into development plans which relate to new development areas.