1The magic dust of quantitative indicators has found its way into policy documents under different banners and logos, such as “sustainability indicators”, “quality of life indicators” and “performance indicators”, at all spatial levels (Sawicki, 2002; Swain and Hollar, 2003). This global trend towards indicator usage resembles the “social indicators movement” witnessed in the 1960s and 1970s when American social scientists, welfare advocates and civil servants used indicators to measure social change and the idea spread rapidly to international organisations such as the Organisation for Economic Cooperation and Development and the Social and Economic Council of the United Nations (Horn, 1993). The current bandwagon of worldwide indicator endeavour began in the 1990s and is named as the “community indicators movement” by Innes and Booher (2000: 174).

2As observed by Wong (2003; 2006), the earlier social indicators movement was very much developed in the context of social reform and welfare at the national level, whereas the current resurgence of interest in indicators has been largely stimulated by broad environmental concerns related to creating sustainability and quality of life indicators at all spatial scales (UNCED, 1992). Following the collapse of communism in the 1980s, the mood for top-down social reform has been less enthusiastic. Instead, it has been replaced by another form of ideology-achieving sustainable development and inter-and intra-generation equality in terms of environmental resource consumption. This new global agenda has not only brought with it a need to employ indicators as a key mechanism for assessing environmental impact and capacity (Maclaren, 1996; Macnaghten et al., 1995), but has also spurred local action and broadened concern to encompass the wider community based issues.

3Since the neo-liberal régime under Thatcher sought to replace formalised legal regulation with market mechanisms and emphasise supply-side, free market and competitive relations between firms and between places (Blakely, 1994), it is logical to expect a trend of decentralisation of indicator collection and usage in Britain similar to that observed in the US (Innes and Booher, 2000). What we witnessed in Britain over the last three decade is, however, quite the opposite. Although there has been a continuous process of delegation of power to local and regional actors to carry out monitoring and data collection, it is concomitantly undermined by the ever-strengthening process of centralisation of funding and performance control (Wong, 2000; 2006).

4While social indicators and sustainability indicators represent the pursuit of some higher order goals and values, there are other less exciting indicators developed by bureaucrats and statisticians to provide day to day management and monitoring of policy regimes. It is in this regard that indicators have never left British policy circles and they were used to inform funding allocation and decision-making for all sorts of urban policy programmes throughout the 1980s. After this introduction, the next section highlights the changing ethos of the British government over the use of quantitative indicators to allocate regeneration resources and to monitor the performance of initiatives. This is then followed by a discussion of the parallel conceptual and methodological development of indicators research in the urban policy field. The Town and City Indicators Framework developed for the monitoring of the Urban White Paper is used as a case study to examine how indicators can be analysed to enhance our understanding of urban change. Some thoughts on the use of indicators in urban policy-making and the main issues to be addressed in future research are then given at the end of the chapter.

Changing policy ethos of indicators usage in Britain

5Indicators such as employment, unemployment and population changes have long been used for urban and regional policy targeting and evaluation in Britain. These measures were employed in the 1950s to identify the differences between development areas and non-assisted areas. The widespread public debate on indicator measurement and usage, however, came later after the publication of the 1981 Deprivation Index. In 1983 the Inner Cities’ Directorate of the Department of Environment (DoE, 1983) published a paper Urban Deprivation which provided analysis to assess the relative levels of deprivation in English local authorities. The 1981 Deprivation Index consisted of eight indicators from the 1981 Population Census. Local authorities that scored within the worst 50 on at least two of the indicators were deemed as the most deprived areas. The rankings of these eight indicators were then used to identify Urban Programme Areas for the eligibility of various physical regeneration grants (DoE, 1993: para. 2). However, the technical exercise of index creation had turned into serious political debate after the government (HM Government, 1986) proposed to allocate rate support grant to local government on the basis of the deprivation indices.

6Intense debate was then started on the conceptualisation of needs and deprivation, the choice of individual indicators, the methods and effects of statistical transformations and the weighting schemes applied to develop the aggregate index. The comments made on the methodology and the choice of indicators by local authorities were all valid and genuine, though it is notable that those authorities that put in a lot of effort to analyse the index were those that had missed out from inclusion in the top 10 most deprived areas, such as Liverpool (Flynn, 1986; Hayes, 1986) and Newcastle upon Tyne (Newcastle upon Tyne City Council, 1986). At the height of the Urban Programme, over 10,000 projects and a total of £236 million (1992-93 figures) were funded each year in the 57 Urban Programme Areas (Cullingworth and Nadin, 2002: 297). This means that significant amounts of public resources are changing hands due to one single technical index. It thus becomes very important that such an index should be subject to more rigorous public debate and scrutiny. This is the main reason why the DoE and its successors’ deprivation indices have continued to be a contentious issue of public debate. Besides the Deprivation Index, the top-down funding allocation framework was also found in other policy areas such as regional assistance and housing capital grant.

7Ever since Mrs. Thatcher came to power at the end of the 1970s, there has been a strong desire to involve the private sector in the urban policymaking process (Hambleton and Thomas, 1995) to replace much of the socialist welfare State with market mechanisms. The emergence of a global economy where there is no spatial constraint on the flow of factors of production, commodities, investment finance and information, poses an increasing threat of international competition for capital investment and market share and led to a belief within the Conservative government that it was up to localises and regions to promote their development, but not for government to redistribute jobs across the country. The market-led enterprise ethos was injected to a whole array of regeneration initiatives by giving local authorities a more strategic role to draw up programmes of action to bid for regeneration funding. The allocation of regeneration resources through a competitive bidding process had started with the introduction of Estate Action in 1986 and City Challenge in 1991. This funding game was subsequently extended in the 1990s to a further suite of programmes including the Single Regeneration Budget, Regional Challenge, Local Challenge and Capital Challenge.

8With the tightening of public expenditure, central government also began scrutinising public programmes and monitoring the effectiveness of individual policy activities more closely. Spatial targeting and co-ordination of European aid also became more important, with the move towards Integrated Development Operation Programmes (CEC, 1991). This led to a whole area of policy activity on performance measurement to demonstrate value for money. As admitted in a European Commission Working Paper (EC, 2000a: 3), the monitoring, control and evaluation procedures of the Structural Funds had been largely focused on financial monitoring. In reviewing the history of urban policy evaluation, Cameron (1990) commented that the Thatcher government made significant progress in urban policy evaluation, though with a tendency of using a restrictive number of performance measures such as job creation or retention rather than the broader impact and outcome of the policy. This criticism was widely endorsed by other commentators (Burton and Boddy, 1995; National Audit Office, 1990). With the pressure imposed by central government, the auditing and performance ethos was also filtering to the management of service delivery in local authorities, though the extent to which different authorities had developed performance indicators varied considerably (Fenwick, 1992).

9Despite changes in the central administration in 1996, the market-driven ideology continued to underpin the policy delivery of Blair’s Labour government as well as influencing the policy framework of the current Brown government. There is a continuons trend of shifting data collection responsibility from the centre to the local, but under strong central guidance. With regard to urban regeneration funding allocation, political signals from the then newly-elected government in 1997 suggested that the principle of competitive-allocated resources to coalitions led by local authorities was likely to remain in place. However, there was a socialist resonance of taking a greater account of needs when determining such resource allocations. The competitive ethos continues to be evident in the housing investment allocation. While the inter-regional allocation of housing capital is carried out on a pure formulaic basis, the approach towards intra-regional is another story (DETR, 2000a). About half of the regional resources was distributed to local areas on the relative needs indices, the other half was at the discretion of the government. However, the allocation of Housing Corporation investment shifted away totally from a formulaic basis to full discretionary allocations in 2002/03. As expected, the discretionary allocation is carried out on a semi-competitive basis by making assessments on a suite of performance indicators.

10At the turn of the millennium, there was a strong push on using evidence to support policy choices and to generate public debate, especially in relation to the key factors and the way government policy affects outcomes (AIU, 2003). This was built upon the experience of the National Health Service to apply research evidence in policy, management and practice decisions (Blackman, 1998). The expectation of the government of rolling out this practice is that “more new ideas, more willingness to question inherited ways of doing things, better use of evidence and research in policy making and better focus on policies that will deliver long-term goals” (HM Government, 1999: para. 6). In order to kick start a culture shift in policy-making, the Performance and Innovation Unit of the Cabinet Office published the Adding It Up report (PIU, 2000) to set out a comprehensive programme for creating the conditions in which rigorous analysis is routinely delivered and demanded for policy-making.

11The publication of the Modernising Government White Paper (HM Government, 1999) marked the government’s strong commitment to performance measures and evidence based policy ethos. The document was filled with buzzwords of performance, targets, objectives, audit and measures. In order to improve public service delivery and performance, Public Service Agreements (PSA) targets were set for improving public services with the aim to shift policy focus from inputs to the outcomes that matter. In addition, Best Value indicators, a locally defined monitoring System of service delivery, were used to underpin performance (DETR, 1999). The important signal projected from the government’s monitoring guidelines was the increasing emphasis on the longer-term horizon of outcome and impact measurement (DTLR, 2002; HM Government, 1999; ODPM, 2002a, 2003, 2005). This coincided with the latest guidelines issued by the European Commission (EC, 2000a) over the monitoring of the new programmes of structural assistance. There seemed to be a realisation that the focus on monitoring outputs in the late 1980s and early 1990s failed to provide a clear picture of policy achievement and there was a need to expand the scope of evaluation to allow for the monitoring of trends and changes (EC, 2000a; SEU, 2000). The recent publication of the 2006 Local Government White Paper (HM Government, 2007) further proposed a radical reduction of national performance indicators from between 600 and 1,200 to a set of fewer than 200, and that they should focus on outcomes rather than inputs, outputs or processes to replace the Best Value indicators in April 2008 (CLG, 2007).

Tracking urban changes: ideas and methods

12Supra-national organisations across the world have shown their interest in assessing urban renaissance across different nations since the millennium. The World Bank (2003) compiles the annual world development indicators series to monitor the achievement towards international development goals. Eurostat has redeveloped “Urban Audits” to improve the knowledge of quality of life in urban areas across Europe (EC, 2000b). A similar venture was carried out by the Asian Development Bank to compile the Cities Data Book (Westfall and de Villa, 2001). Indicators on urban and regional development tend to appeal to a wide range of stakeholders and are frequently under the limelight of the media. This is because indicators measuring characteristics of places tend to find their way into place ratings and league tables. The chasm between traditionally highbrow academic research and the popular public and business interests can thus be bridged by such statistics.

13The perplexing issues involved in the development of urban development indicators mean that they do not always produce what they appear to promise (Wong, 2003). Under the neo-liberal ideology of the British government, and its advocacy of a strong auditing culture, a range of indicator studies were commissioned by different government departments to measure their specific policy concerns (see examples in Table 1). The nature of these official studies tended to be short-term and ad hoc. This meant the scope for fundamental theoretical and methodological analysis was highly constrained. For instance, the conceptualisation of key issues like “competitive advantage” and “economic performance” in a report prepared by PA Cambridge Economie Consultants (1990) appears to be rather superficial. Pieda’s (1995) practical guidance on local economic audits focuses on the steps and procedures of measurement, but key issues and factors included in the proposed framework were not thoroughly explained. In addition, some of this commissioned work was in the form of pilot studies which did not involve the compilation of a complete data set (e.g. Coombes, Raybould and Wong, 1992; LGMB, 1995).

14The growing need to identify urban regeneration potential and needs, and increasing policy interest in statistical information, helped spur significant academic interest in measuring socio-economic conditions and prosperity.

15These include the “booming towns” study by Green and Champion (1991), Breheny et al. (1987) “northern lights” research, the “quality of life” study by Rogerson et al. (1989), and the measurement of geodemographics by Brown and Batey (1994). These research studies have rejuvenated academic discussion over various methodological issues, for example, ways of quantifying intangible issues (e.g. Rogerson et al., 1989), the derivation of appropriate weighting Systems to create composite indices (e.g. Green and Champion, 1991; Rogerson et al, 1987), and use innovative methods to develop area classifications (e.g. Brown, 1989; Brown and Batey, 1994). On the whole, these academic studies tended to produce place-rating schemes based on economic performance or geodemographic profile, rather than bearing any specific policy concerns in mind. The outcome of the resulting league tables was often controversial, partly because they were subject to misinterpretation by the media. However, such media hype on city development league tables helped to raise the attention of policy makers, as well as furthering academic interest, in quantitative measurement.

16The nature of this wave of urban indicators research in Britain (see Wong, 2003), both policy-related and pure academic studies, tends to be strongly grounded in an empirical approach. More attention has thus been paid to the methods of measurement and the empirical exploration of data. In spite of the shortcomings and constraints of commissioned studies, new methodologies and research insights have spun off from earlier consultancy research.

Monitoring urban renaissance: the Town and City Indicators Framework

17In England, the Urban White Paper (DETR, 2000b) sets out an agenda for delivering an urban renaissance for towns and cities in England. It explicitly identifies five visions to provide a better quality of life for everyone in our towns and cities:

• People shaping the future of their community, supported by strong and truly representative local leaders.

• People living in attractive, well-kept towns and cities which use space and buildings well.

• Good design and planning which makes it practical to live in a more environmentally sustainable way, with less noise, pollution and traffic congestion.

• Towns and cities able to create and share prosperity, investing to help all their citizens reach their full potential.

• Good quality services-health, education, housing, transport, finance, shopping, leisure and protection from crime-that meet the needs of people and businesses wherever they are.

18In order to better understand how to monitor urban change and track progress of policies towards achieving these visions, a Town and City Indicators Database (TCID) (see Wong et al., 2004) was developed as part of a long-term evaluation framework towards a wider urban policy evaluation strategy and to provide input to the preparation of the State of Cities Report planned for 2005. Unlike the earlier research discussed above, this study focuses on methodological development as well as the innovative analysis to highlight key trends in urban change across England.

Conceptualising urban change

19Urban change is a continuous process of spatial transformation, which includes economic, social, physical and demographic dimensions. The five visions of the Urban White Paper encapsulate these elements of change. The driving force can be due to structural changes and historic inertia at the local level as well as external factors from national and global forces, or indeed the interaction of internal and external factors (see Figure 1). Global financial flows and the use of telecommunications technologies have altered the international economic development landscape in which cities compete. At the same time, long-standing economic decline and social problems in urban areas have triggered labour market restructuring and other socio-cultural adjustments. These changes transcend the neatly defined administrative boundaries of local authorities. The outcomes of the process of change have been mixed and are contingent upon the endowment and exploitation of assets and resources of the urban areas and their wider functional hinterlands. Greater understanding of the dynamic interactions of different activities at different levels of our urban areas will help inform the delivery of urban policy. The development of the TCID aims to monitor these key trends of urban change.

20The terms “town” and “city” are used to refer to “physically built-up” urban areas. For largely practical reasons of data availability and relevance, urban areas are defined as physically built-up and having a population of more than 20,000. A small number of these urban areas have been subdivided because they include separate towns with few interconnections. Government Offices in all regions were consulted on the principles of these definitions and were supportive. There are altogether 257 primary urban areas (PUAs) in the database, although some indicators are only available for a sub-set of 78 PUAs.

Research strategy

21A multi-faceted research strategy was adopted to develop the TCID and analyse the baseline position of towns and cities. First of all, a two-tier indicator System is used to underpin the overall structure of the indicator framework. This means that there is a separation of the more strategic overarching indicators of urban change at the top tier and domain-based vision indicators for the five urban visions at the lower tier. Strategic indicators are used to encapsulate trend data on a small number of indicators that are widely used by researchers to gauge urban change. Trend data are collected to provide a picture of the condition of urban areas brought about by the process of socio-economic restructuring. These indicators measure both intensity and dynamics. Examples include population level and change; employment level and change; and unemployment level and change and duration. The lower tier of the indicator System focuses on dealing with domain-based issues guided by the underlying conceptual framework of the five Urban White Paper visions.

22The study also adopts a structure-performance model to highlight the differential socio-economic contexts against which urban areas operate, so as to provide a sound basis for interpreting performance and change on a like for like basis. The model used is based on classifications of PU As by size, regional location and on a shift-share categorisation of employment change. Shift-share analysis is essentially a means of breaking down an area’s total employment change into components representing a set of influences: the national employment trend, employment trends by industry and the relative mix of industries in an area and a “residual”, representing the effect of other location-specific factors upon employment change. The shift-share analysis component thus provides a summary of the changing employment structure of towns and cities, which is one of the primary drivers of the structure-performance model.

23The third strategy in the research design is about the focus of PUAs as the main spatial unit of analysis. Local authority districts are used as the building block for a majority of indicators, as most datasets are available at this level. Efforts are, however, now being made to move away from using administrative boundaries and to derive more data for micro areas, so as to produce data for accurately defined urban areas. The development of Super Output Areas by Neighbourhood Statistics offers the prospect of aggregating data from micro areas to fit the physical reality of towns and cities.

24With regard to the strategy of analysis, the research adopts the analytical bundle approach. There are alternative ways of simplifying the structure of the indicator set including the use of composite indices. However, the technical approaches of applying weightings to develop composite indices tend to be either arbitrary or too complicated to be transparent. This study, therefore, adopts a more direct, analytical approach by grouping indicators that measure a particular dimension of the urban vision into “bundles” to provide commentaries on the spatial patterns emerging from these indicator values of the five urban visions.

25In order to ensure that the TCID and associated analyses provide useful information to inform urban policy, six local authorities1 were interviewed to seek their comments on the data collection and analysis framework. With regard to the development of the database, a total of 125 indicators for the five visions set out in the Urban White Paper were identified and assessed by the research team. The assessment of indicators was carried out systematically based on a standard set of criteria: accessibility of data sources, input and output spatial units, the date to which the information refers, prospect of updating, availability of a time series, storage format, ease of implementation and interpretation, and relevance to the urban vision objectives.

An overview of urban change

26A strategic overview of the main features of urban change was ascertained through a limited number of strategic indicators on population change, employment, unemployment and GDP.

27The trend over the period from 1991 to 2001 was for greater relative growth in population at lower levels of the urban hierarchy alongside population loss in the metropolitan areas outside London. Migration acts as the key driving force to increase population at lower levels of the urban hierarchy and to reduce the population in metropolitan areas outside London. Since the mid-1980s, the prevailing trend of counter-urbanisation of population has been reversed at the very top of the urban hierarchy, with a return to population increases in London (and evidence of this tendency emerging in other major cities). Over this period, London has witnessed a substantial increase in population, mainly as a result of international migration. London has by far the largest concentration of population from minority ethnic groups of any of the area categories. The fertility of people from minority ethnic groups and of people born outside the UK is higher than the UK-born population, and London’s population growth is now also being driven by births to recent international migrants. Other large urban areas also display greater than average concentrations of population from minority ethnic groups, which also displayed higher than average rates of population increase during the 1990s.

28An increase in employment was recorded throughout the urban and regional System between 1991 and 2001, but relative gains in metropolitan areas outside London, and in freestanding cities and industrial areas, were less pronounced than the national average. The largest percentage increases in employees were recorded in inner London and in mixed urban-rural areas. The latter category records the highest employment rates in England, while the largest urban areas record the lowest employment rates. There was a marked decrease in unemployment over the ten years from 1991. Despite the large reductions in the number of claimants and in the unemployment rate over the period from 1991 to 2001, unemployment remains disproportionately concentrated in the large urban areas.

29GDP per head is higher in inner London than in any other part of England, and the trend over recent years has been for a widening of the “gap” between London and mixed urban-rural areas on the one hand, and the rest. Lowest levels of GDP per head are recorded in other metropolitan authorities, resorts and industrial areas.

30Overall, London has seen greater than average gains in population, employment and GDP. The fortunes of the other largest cities and their associated metropolitan areas have, nevertheless, been less favourable as indicated by continuing population loss and smaller than average employment gains. Unemployment and non-employment remains concentrated in the largest urban areas in England (including London). The mixed urban-rural areas have witnessed the most favourable changes on most strategic indicators.

Performance of towns and cities

31After a comprehensive assessment of potential indicators, only 90 were recommended for inclusion in the TCID. The decision to drop the other indicators is attributable to various reasons, including the lack of appropriate data sources, the indicator being too ambiguous for interpretation and/or the indicator being of only minor relevance to the urban visions. There is a paucity of robust indicators for Vision 1 (People shaping the future) and Vision 5 (Good quality services), and less than half of the indicators can be measured at more fine-grained spatial levels to allow a best-fit aggregation to all 257 PUAs. Analysis was First carried out for all 90 indicators, and after further evaluation, only 55 (listed in Table 2) were deemed as reliable and provide meaningful interpretation of urban issues at the time of the research.

32Analysis of the 55 vision indicators was carried out for London and the eight Core Cities-Central West Midlands (where Birmingham is the largest city), Bristol, Leeds, Liverpool, Manchester, Tyneside (where Newcastle upon Tyne is based), Nottingham and Sheffield-to compare and contrast the baseline position of the largest cities in England over different aspects of the State of urban change. The analysis of these vision indicators portrays a very complex picture of urban change in towns and cities across England. A number of key patterns of development, however, can be identified.

33There are differentials in the experience of urban change in different regions. These tend to demarcate the contrast between the South East, London and the East of England and the rest of the country. The three northern regions, the North East, Yorkshire and the Humber and the North West, tend to display patterns akin to each other and face similar issues of urban change. These broad patterns of urban development are highlighted by indicators on population change, housing issues, deprivation, employment and unemployment patterns. Although the broad patterns of development appear to feature a “North-South divide”, the dividing line is not that watertight. The South West, the East Midlands and West Midlands are akin to Southern regions on some indicators, and to northern regions on others. London itself is a region that has a very distinctive pattern of development from the rest of England. Moreover, it is also important to point out that within each region, the patterns of development among urban areas vary substantially. As shown from the shift-share analysis, those areas enjoying more favourable conditions tend to out-perform those urban areas operating in less favourable conditions in terms of growth in employment and economic activities.

34The analysis of the vision indicators also lends support to the broad patterns of spatial decentralisation from large urban areas to smaller ones. The smaller urban size groups, especially small cities (e.g. Southampton, Bradford and Reading), large towns (e.g. Wakefield, Peterborough and York) and small towns (e.g. Taunton, Barrow-in-Furness and Whitstable), tend to have experienced more favourable change than larger urban areas in terms of population change, employment growth, proportion of qualified workforce, new build housing stock and planning applications. Nevertheless, large urban areas have contributed more towards achieving a sustainable environment in terms of having higher dwelling densities and using brownfield land.

35London stands out as a major economic powerhouse over a whole array of indicators in comparison with the Core Cities. However, it also exhibits high house prices, poor air quality, unsustainable commuting patterns and unemployment problems. The patterns of development in the main cities are mixed. On the whole, Bristol and London, and to a large extent Nottingham, display more favourable development patterns than the other Core Cities. However, there are signs of positive urban change in Leeds, Manchester, the Central West Midlands and Sheffield.

***

36The analysis of the TCID project demonstrates that it is feasible to collect and analyse a large number of indicators to track changes in urban areas. There are, however, a number of valuable lessons to be learnt from the complex and time-consuming data compilation and analytical exercise.

37The focus of this research study was on establishing the baseline position of towns and cities around 2001 to provide a platform for future trend analysis. This meant that the scope to examine urban change at this stage was limited and was largely achieved through the trend analysis of strategic indicators and the dynamic vision indicators. It is, however, important to point out that any meaningful trend analysis will be jeopardized by changing variable definitions and data compilation methods. The experience from this project shows that the methods of collecting and compiling various national statistics have been subject to constant revisions and adjustment, such that during the course of the project it was necessary to recalculate some indicators more than twice.

38In spite of the fact that the TCID aims to measure the State of urban areas across urban renaissance visions, there is a need to have more descriptive contextual information on different urban areas to provide a fuller backdrop to the trend analysis. Furthermore, the analysis of the conditions of a particular urban area can only be enhanced through benchmarking-by region, urban size group and shift-share category. On the whole, the region and urban size group classifications provide the most useful insights. There are certain practical difficulties in applying the shift-share categories in the analysis. Shift-share analysis is a descriptive, data-partitioning technique, but its results have sometimes been given too great theoretical significance. The output of the technique is extremely sensitive to the quality of the input data, in particular with relation to the sample design and reclassification of industries over time.

39The vision indicators collected in the TCID are not evenly distributed across the five visions. It is difficult to find appropriate indicators to measure certain issues such as community participation, aesthetic quality and attractiveness of towns and cities, quality of transport System, and health and life satisfaction. While a number of survey-based Best Value indicators were collected and analysed in the first instance, the final decision was that they did not provide reliable and meaningful interpretation of the issues concerned. This leads to the conclusion that there are certain issues which are simply not susceptible to quantification, and hence that other forms of qualitative information should be collected to provide policy intelligence on such issues. However, these survey-oriented data could only be used if concerted effort were made to develop robust sampling methodologies and to identify valid questions to improve the quality and coverage of such surveys.

40A large amount of data from a whole variety of sources were collected and assembled centrally to form a Consolidated database for this project. As far as possible, the data were collected on a common spatial and temporal basis. However, less than half of the indicators are available at sufficiently fine-grained spatial levels (i.e. wards and postcode sectors) to allow a best-fit aggregation to all 257 PUAs. The remaining indicators, by and large, were calculated using datasets available at the unitary authority/local authority district level. This means that only a sub-set of 78 PUAs can be best fitted with these datasets to produce meaningful measures. This situation can only be improved when better small area data and neighbourhood statistics become available. The Census of Population remains a crucial source of information for analysis at the micro area level. However, the key disadvantage of the Census information is that it is available only on a decennial basis. During the process of compiling the TCID, some administrative databases were found containing information that could be reported at more fine-grained scales. However, due to the need to preserve confidentiality, these datasets tend to have restricted access. This means that special arrangements may have to be negotiated with the relevant data holders in order to maximise their potential for monitoring urban condition and change.

41The TCID framework addresses the long-standing challenges of measuring the effectiveness and outcomes of urban visions. While accepting that no single set of indicators will ever be optimal and perfect, the solution is to use different research strategies to improve the analysis of indicators to yield meaningful policy intelligence. The indicators selection framework and the analytical principles set out in the TCID project should provide a platform to allow effective monitoring of urban strategies over time. It is important to note that the identified indicators can change if new policy issues emerge and better quality datasets come along. At the beginning of 2003, the government published its Sustainable Communities: Building for the Future document (ODPM, 2003), which set out a framework of action and investment to address its priorities in relation to housing demand and supply, low demand, empty homes, the use of land, liveability and housing conditions. The concept of sustainable communities very much echoes the five urban renaissance visions set out in the Urban White Paper. More recently, the publication of the Local Government White Paper in 2006 has shifted the discourse to the place-shaping and liveability agenda. In spite of the changing discourses, it is clear that the issues covered and indicators collected for the TCID could play a key role to monitor progress and development towards the achievement of sustainable and liveable communities. Under the evidence-base policy regime, it is anticipated that there will be a need to rationalise and harmonise different indicator sets to avoid information overload and secure effective policy application. The radical reduction of local government performance measures to 198 indicators can be seen as the beginning of this bigger indicator harmonisation project.