(C²I)² = CCI-CCI The CCI Creative City Index 2012

Chapter 1 Why a City Index Matters

Chapter 2 Critical Review of the City Index Industry

Chapter 3 Elements of the CCI Creative City Index (CCI-CCI)

Chapter 4 The CCI Creative City Index and Results

Chapter 5 Conclusions

The CCI Creative City Index (CCI-CCI) is a new approach to the measurement and ranking of creative global cities. It is constructed over eight principal dimensions, each with multiple distinct elements. Some of these dimensions are familiar from other global city indexes, such as the MORI or GaWC indexes, which account for the size of creative industries, the scale of cultural amenities, or the flows of creative people and global connectedness. In addition to these indicators, the CCI-CCI contributes several new dimensions. These measure the demand side of creative participation, the attention economy, user-created content, and the productivity of socially networked consumers.

Global creative cities can often seem alike, in respect of per-capita measures of factors such as public spending on cultural amenities, or the number of hotels and restaurants. This is to be expected when people and capital are relatively free to move, and where economic and political institutions are broadly comparable. However, we find that different cities can register far larger differences at the level of consumer-co-creation and especially digital creative ‘microproductivity’. To explain this finding, we review the logic and rationale of creative and global city index construction and present a review of previous and contemporary indexes.

We set out the case for our new model of a creative city index by showing why greater attention to consumer co-creation and microproductivity are important, as well as examining how these factors have been previously overlooked. We show how we have measured these additional factors and indicate the effect they have on creative and global city indexes.

We then present the findings from a pilot study of six cities, two Australian, two German and two from the UK, to indicate how the new index is calculated and applied. Our results indicate much greater variance arising from the new arguments between cities.

Competitive global cities

A key insight that all city indexes point to, as ours most certainly does, is that creative global cities are increasingly engaged in intense competition with each other in the evolving process of globalization.

Globalization is a centuries-long process that results in the increased interdependence of peoples’ economic, cultural, social and political lives. Globalization progresses as ‘factor mobility’ increases, and accelerates when people, capital, money and ideas are free to move about the world to settle where they are most valued. In the past few decades the world has experienced its greatest wave of globalization. Mobility is at unprecedented levels.

From 2008 a majority of the world’s population – over 3 billion people – now lives in cities, making humanity an urban species for the first time (UNFPA 2007). In completely new ways and at unprecedented scale, human experience is city life, a reality to which current thinking and policy settings have not yet adjusted:

Mobility, urbanisation, and technology have converged on the contemporary city, which, although fixed in place, is best analysed as a dynamic hub in a global network. Every place is now connected to every other place, and so, therefore, is every person. Interestingly, that means that cities are now the most important unit of social-cultural and economic organization.

It also means that cities compete globally. Nation-states are no longer the key units of global competition; instead, cities compete with one another for valuable scarce factors of production. The most important factor, by far, is ‘human capital’ – enterprising, talented and creative individuals. The focus of modern globalization is creative people; and creative cities are the product of their interactions, driving socio-cultural and economic evolution.

Competing city indexes

That globalization is a selection mechanism on cities is the key insight that underpins contemporary attention to the measure of a city’s creative output and potential. The demand for and development of such indexes has burgeoned in the past decade (see chapter 2 of this report). Yet a central challenge remains: what is the effective measure of a creative city? This has turned out to be a surprisingly difficult question to answer. It is not simply a matter of adding up a city’s capital infrastructure and knowledge exports, or adducing a measure from population size, although these factors are important. Degrees of openness and diversity also need to be included in any measure of a global creative city. These have been key additions contributed by the most influential players, the Florida, MORI Foundation and GaWC indexes. The current state of the art is that manufacturing-based indexes have been updated to account for the openness, global integration, attractiveness and liveability of a city to produce much improved measures of global city indexes and rankings.

Yet there remain significant gaps in this enterprise, most notably in accounting for consumer imagination, user co-creation and amateur production, and the social learning dynamics of the creative citizen, who is connected both to ‘small world’ networks (see e.g. Ormerod & Wiltshire 2008) and to global complex systems via digital social networking, including digital platforms like Facebook (Potts et al 2008a). These factors remain unaccounted in current indexes for at least two reasons: (1) they are difficult and seemingly subjective measures to make; and (2) in traditional industrial economics there was little reason to suppose that these factors mattered.

Our work at the CCI, particularly through the intellectual framework developed in John Hartley’s Federation Fellowship program (e.g. Hartley 2009; 2012), has highlighted the increasing importance of what we call here ‘microproductivity’ in the cultural, social and economic impact of the creative industries. The work of Jason Potts on the drivers of economic evolution has reinforced this observation, which further recognises the ‘situatedness’ of productive creativity (Potts et al 2008b; Potts 2011).

Between them, these approaches indicate that ‘consumer’ activities, the non-professional creative productivity of ordinary citizens in digitally linked social networks, should be regarded as part of the innovation system of complex cultural economies; and that the paradigm example of such systems is the contemporary world city.

Cities are crucibles of everyday human inventiveness through the rapid experimentation, market feedback and social copying processes that drive creative endeavour. Some cities do this better than others, and those that do can become great creative cities.

But how do we recognise this ‘on the fly’, as the processes unfold, and as novelties emerge? How can we identify when and how a city is doing well, even as it is doing so? That is the purpose of a creative city index. This report will present an analysis of why we need a better index model, and a working prototype of such an index.

The challenge lies in selecting a set of indicators to feed critical information to a dashboard index, including one that can measure a city’s absolute performance (how creative?) and its relative performance (compared with which other cities?). The report details our endeavour to construct such an index from first principles, along with the rationale for why we have included the indicators that we have.

Most important, we present a worked example of the index applied to six cities – two each from Australia (Brisbane & Melbourne), Germany (Bremen & Berlin) and the UK (Cardiff & London). We chose London because it regularly features in the top three of any index of global cities (along with New York and Tokyo); but by the same token it is far from typical. Thus, we paired it with Cardiff, a more compact city that is also a capital and has its own claims to creativity. We wanted to compare different countries, not least because each has its own peculiarities in the collection of data and compilation of statistics, not always in English. So we have chosen a pair of cities – one metropolitan and one provincial – in each of three countries, not all of them Anglophone: Australia; Germany; the UK.

Our index is calculated and presented as unweighted, in that the summation weights given to each of the 8 index components are equal. That is because we have no priors relating to the relative importance of each component to add to this. Yet even without such calibration, we propose that its raw form offers a superior measure of the creativity of a city compared with other indexes, by assembling not only its industrial output and its cultural attractiveness – these are the two factors that dominate extant indexes – but also the contribution of its creative citizens. We believe this third impulse to be the main motive and driving causal factor in creative city development.

Our index is built on an approach that can be dubbed ‘cultural science’.

See http://cultural-science.org.

We show how we have measured these additional arguments and indicate the effect they have on creative and global city indexes. We then overview the findings from a pilot study of six cities (two Australian, two German and two from the UK) to indicate how the new index is applied. Our results indicate much greater variance arising from the new arguments between cities.

Our findings are summarized in the following table and figures.

Figure 1

Figure 2

In September 2010 the Beijing Research Centre for Science of Science (BJSS),

CCI: www.cci.edu.au; BJSS: www.bjss.org.cn.

The work has been carried out by Prof. Hartley, Dr Jason Potts (CCI Centre Fellow) and Trent MacDonald (CCI Research Associate), with the assistance of BJSS researcher and CCI postgraduate Angela Lin Huang. Research assistance for the population of the indicator suites was provided by Chris Erkunt and Carl Kufleitner of the University of Queensland.

Figure 3

The report surveys the modern development of city rank indexes and outlines the method and theory for the construction of an improved city index that identifies creative (world) cities. The aim of the report is to produce a comparative assessment of city indexes and rankings, showing what data have been used to generate each index, and highlighting the extent of creativity and cultural indicators. We outline our initial construction of an alternative suite of indicators and index construction methods.

We have sought to produce a ‘Creative City Index toolkit’ for the municipal government of Beijing to use in domestic and international creative city benchmarking and comparison, thereby to promote policy settings that will cultivate a creative and cultural environment of world-class standing, with the strategic aim of transforming Beijing into a leading global creative city. The CCI Creative City Index (CCI-CCI) is designed to measure and track a city’s creativity over time, to make comparisons with peer and exemplary cities worldwide, and to provide a mechanism whereby cities can be internationally ranked.

The index will be used as follows:

Beyond this immediate purpose, the CCI-CCI indexing tool can be adapted on demand to provide a creativity score and ranking for any city. Its further development outside China will be the responsibility of the CCI.

The work underlying the development of a Creative City Index poses important new questions: What is creativity? What is the creative sector of the economy? What is a creative city? And what is the creative contribution of ordinary citizens? These questions form the context for this report. The CCI Creative City Index toolkit is framed by the recent conception of creativity and the creative city developed by the CCI (QUT) while also being nested within the extant creative city literature.

According to the brief given to the CCI by the BJSS, the scope of this project is as follows:

The BJSS will perform data collection work and compile the CCI Creative City index for Beijing and other cities inside China, for comparison, if so desired. For cities outside China, the CCI-CCI will be developed by the CCI.

What makes a great city? This is the question that the construction of world city or global city indexes has sought to answer. Interestingly, this concern has been a very recent development, mostly of the past decade, broadly coinciding with the rise of globalisation, the emergence of rapidly growing and changing emerging economy cities, such as Shanghai, Beijing, Hong Kong, Tokyo, Mumbai, etc., that are increasingly in direct competition with ‘old’ global cities like New York, London, Paris, etc., and coupled with vast improvements in information collection and classification standardisation.

Great Cities 1 – Power

It was not so long ago that the answer to what makes a great city was a simple arithmetic of:

Population ×<!-- × --> Wealth + Empire = Power $$\text { Population } \times \text { Wealth }+\text { Empire }=\text { Power }$$

New York, London and Paris were big trading cities, capitals of empire, and rich: that made them powerful global cities without much further consideration of the specific causal components. There were certainly bigger cities (Mexico City or Jakarta, for example), but they were poorer, and thus plainly not global cities in the modern sense. Earlier European power-cities like Florence, Venice, Genoa, Lisbon, Amsterdam and Vienna would successively have figured on the list; as would capitals of non-Western empires, such as Istanbul. Going back further we would include Rome, Constantinople, Alexandria and Athens; and further still, Damascus, Byblos, Luoyang, Xi’an or Varanasi/Benares.

And see: http://en.wikipedia.org/wiki/List_of_cities_by_time_of_continuous_habitation.

City historians Peter Hall (1998) and Joel Kotkin (2005) provide scholarly overviews of the historical approach to city research. For more detailed statistical treatment, the work of Michael Batty (2006) on city scaling laws uses data on global city populations over the past 3000 years to trace these changes. There is no shortage of thoroughly researched long-run accounts of the rise and fall of great cities. For the most part they emphasise a simple political-economic arithmetic of a locus of power combined with fertile plains and effective institutions. Great cities were capital cities; places where power resided.

By the Middle Ages, cities became places where commerce resided too, as power began to shift to the secular world. Much of this story is associated with the rise of the West (Ferguson 2011), the invention of science, technology, democracy, media and, most important, the institutions of trade and commerce. Here global cities began to emerge about trading ports and network hubs (e.g. the northern European Hanseatic League), not about castles and palaces. Thus the world began to urbanise as people came to cities for reasons of enterprise, not for alms (or arms) or rents.

To this day, cities may still be plausibly divided by this distinction (patronage vs. entrepreneurship), a distinction that as yet has not been measured.

Great Cities 2 – Enterprise

A somewhat different line has also developed through the work of urban theorists such as Jane Jacobs (1969) and more recently by Richard Florida (2002; 2005) and Elizabeth Currid (2007). This approach focuses on the adaptive functionality and emergent liveability of cities, emphasising in essence that great cities are grown, not planned. By drawing on the economics of self-organisation, Jacobs shows how – contrary to their purpose – planning laws and great plans often stymie the development of cities, by suffocating the development of local neighbourhoods. Florida and Currid show how the development of a city depends upon who moves there and why, emphasising the mobility dynamics of cities.

This dynamic institutional line has also been explored recently by economists Paul Romer (1990) and Ed Glaeser (2010),

For Paul Romer, see also: www.chartercities.org. On Edward Glaeser's work see www.economics.harvard.edu/faculty/glaeser/papers_glaeser.

In this view, cities are complexes of people, and the more entrepreneurial and capable those people are, the greater the city will be. Furthermore, because of feedback effects that accrue to the concentration and proximity of such idea-carriers and generators (Romer 1990, Krugman 1994), city greatness exhibits non-linear dynamics.

Great Cities Now – Great Attractors

Crucially, this introduces the importance of great cities as zones of attraction. In the older definition, a great city was so because it was powerful and rich; it was the centre (indeed the storehouse) of past greatness. In the new definition, however, a great city is so because those that seek to become rich choose to move there. It is great because of the present and future potential it offers, not because of its past accumulations. In this emergent dynamic conception, a great city is something that can be triggered and can shift in relatively short order as a function of the inflow of smart new people.

This shift in focus from a romantic and mostly static view of great cities as a capital consequence of great empires, towards the modern globalised and entirely dynamic view of great cities as zones of attraction, shifts the analysis of city greatness. In the older model, there was no serious discussion of what makes a city great because the answer was contained in the question of what makes an empire or monarch great. City greatness was an accidental consequence of being the capital of such an empire. There was nothing to understand about city greatness per se, and much to misunderstand.

But as empires have crumbled since World War I, and nation-states have weakened as economic entities since globalisation, cities have re-emerged (as they were prior to the 1700s) as the crucibles of commerce and the proper focus of development and growth economics.

This is evidenced by the fact that the leading growth and development economists at the most elite universities have focused on the causes of city growth, to a first approximation, as the outcome of the choices of a mobile elite of smart global citizens. Great cities, in this view, are made of interesting and capable people. To make a great city, the issue then becomes how to attract such great people.

New scholars of cities, following the line of evolutionary and institutional economics, have re-emphasised the role of good rules or institutions (economic efficacy). The new model of cities is dominated by entrepreneurship and attractiveness to citizens. Cities are great because of their perceived capabilities and potentials. The question, then, is what attracts capable citizens and potential immigrants? The purpose of index construction is to answer that question.

Great Cities Need a Great Index

Recent scholarly endeavour to answer the question has sought to follow these new theoretical lines and has turned to the construction of multi-component statistical indexes, each based around a suite of specific indicators over multiple dimensions. Such indicators seek to decompose and identity the factors that attract interesting migrants and citizens.

This has seen something of a growth industry in city analytics during the past 10 years or so. Richard Florida, for example, has famously emphasised talent, technology and tolerance as the secret ingredients of the mix. Others such as John Kasarda (2011) emphasise the role of growth focused transport networks, or specifically what he calls the ‘aerotropolis’. Despite their differing attention to the role of the ‘creative class’ and that of the airport, these approaches emphasise the same thing, namely the city as an attractor, or – it follows – as a competitor with other cities. In Kasarda’s world, it is not companies that compete, but supply chains. In Florida’s world, it is not people who compete, but cities. This should be well familiar to students of strategic management theorist Michael Porter, who makes an integrated point about the development of localised competitive advantage. The very definition of a (successful) cluster is such a nexus. Economists such as Alfred Marshall had already recognised and described this phenomenon over 100 years ago with the notion of ‘external economies’ which then became the basis for work on industrial districts and clusters.

But now we need to try and disentangle this complex, to understand what specific aspects contribute to this contemporary and ongoing attractiveness. And that is the point of the various global city indexes that have recently emerged. Once we are beyond the notion that great cities are the children of great empires, we enter a world where great cities are the ‘adults’ of great institutions. But what are those institutions? That is what we seek to discover. The purpose of the extant and new city index constructions is to unpack this complex.

The value of these indexes is thereby two-fold. On the one hand they quantify and standardise measures of variation. On the other hand they decompose it into factors of important variation. We are now concerned to seek to understand the specific sources of variation and differentiation that make a great city great. This is no longer a default presumption by capital position in empire.

Cities now compete. And like the competition between corporations in highly contestable markets, these dimensions and magnitudes of differences in cities are issues of significant competitive advantage. We are now in a world where rank indexes of cites matter not as politically charged issues of mostly romantic issues of imperial pride, but as practical issues of city strategy. These indexes are less celebratory laurels of past glory, and more dashboard measures of current performance. These new indexes matter; a good index is a competitive advantage to any growing and competitive city.

There is a substantial body of literature on the rise of global cities. This divides between the work of historians, such as Peter Hall and Joel Kotkin, urban sociologists such as Saskia Sassen, Carl Abbot and Manuel Castells, and urban social geographers such as Paul Knox and Peter Taylor, among many others.

Scottish sociologist and town planner Patrick Geddes, for example, arguably developed this field in Geddes (1911).

First, cities matter. They are the primary locus for the development of human civilisation and economic systems. This point is unambiguous in the work of city historians, and it carries over to the ‘new economic geography’ and ‘endogenous growth theory’, associated with the work of economists Paul Krugman and Paul Romer, for example. Cities are crucibles of economic and socio-political development. The key point here is that this statement thereby excludes or relegates other such levels of organisation, such as villages, towns, nation states, international communities, or even empires, as being relatively less significant. The explanation is the economics of ideas. Cities are places where people and ideas mix most effectively.

Second, cities rise and fall. The rankings of great (world) cities change through time. These changes in part reflect the vicissitudes of other exogenous factors impinging on regions, nation states, empires and the like, but are also due to the endogenous success of particular city institutions. There is mobility in the rank of cities through time, and achieving greatness is no guarantee of maintaining that through time. Again, this is a robust finding from city historians (see Batty 2006, op. cit.).

Third, cities compete. They compete for ‘factors of production’ which means, in effect, people and capital. For as long as factors of production have been mobile, cities have competed. Their success or failure in competition for these globally mobile factors is the major explanation of the rise and fall of cities.

This insight has provoked two types of response among scholars. One side, represented mostly by historians and economists, sees competition as a natural and evolutionary process and emphasises its gains and benefits. It implicitly values the mobility of factors (including people) as a free choice that results in the aggregate improvement of an ever more globalised society. It tends to look for the successful strategies and institutions that have enabled some cities to compete successfully. The other side, represented mostly by urban sociologists and geographers, sees the problems and troubles caused by mobility-induced competition (globalisation as a problem). It emphasises dislocation, social problems, destruction of community and the like. It looks to city-level government intervention, in particular planning-based solutions, to redress these problems. There are of course many exceptions to this binary opposition, for example the anti-planning urban sociologist Jane Jacobs. Partial resolutions are achieved, for example by sociologists turned strategists such as Charles Landry and Richard Florida. But there remains a significant difference in analysis and policy thrust between those who view competition between cities positively, as an opportunity with strategic solutions, and those who view it negatively, as a problem with intervention solutions.

Our report finds greater congruence with the historians’ and economists’ perspective than with that of the sociologists and geographers. We seek to accept all three propositions: that cities matter to human and economic development; that cities rise and fall; and that this dynamic is a consequence of the process of global competition for scarce, valuable and mobile resources, the most important of which are enterprising and creative people.

We recognise that this dynamic evolutionary process will create disruptions and transition problems; including ‘creative destruction’ of city infrastructure, legacy cultural forms and their attendant occupations, and that there will be scope for management and even planning in this process. But we do not conclude from this that the top-down management of downside dynamics (which we call a ‘welfare’ model of cultural policy) should be the main focus for a presently successful or aspiring global city. If the watchword is competition, which plays out globally, then an entrepreneurial and strategic response is required.

When framed in this way, our underlying ‘model’ is one of dynamic flows. A global city is not understood with respect to exogenous factors such as a given political power base, or industrial base, or even cultural base, but rather with respect to the endogenous question of what makes the people and other factors who reside there more productive and effective than they would be elsewhere (i.e. why do existing ‘customers’ stay?) and why do people and other factors who are not there want to migrate there (i.e. why do new ‘customers’ arrive?). This model then focuses on the creative productivity benefit that a city offers, and by implication its supporting or enabling physical, social, legal, institutional and cultural infrastructure, both tangible and intangible.

What competitive advantage (in the language of Michael Porter) does a city’s creative potential generate; and how can that advantage be sustained?

In Charles Leadbeater's terms, a sustainable competitive advantage in the knowledge economy requires creativity that is 'hard to imitate,' producing IP that is 'easy to replicate': see Leadbeater & Oakley (2001), p. 19.

But we must face the problem that ‘attractive factors’ and ‘creative potential for competitive advantage’ are difficult to identity and define. This is why it is necessary to construct a suite of indicators to capture activities that may seem insignificant; spaces for youth culture, for example, which on the face of it are less impressive than great airports or industrial headquarters. Appropriate policies are difficult to implement, because simply committing large amounts of public resources may not be sufficient. It may not even be necessary. Grand plans often fail. As in all competition, adaptation, experimentation and learning are the key process ingredients. Rather than recommending a shopping-list of capital infrastructure investments, the approach we develop is focused on identifying such indicators of experimentation, learning and innovation. This is especially necessary in the challenge of creating a creative world city.

Creative cities of the past did not arise from pre-constructed plans, but emerged as self-organised solutions to problems of complexity, as Jane Jacobs (2000) put it – from an experimental and often stumbling process of adaptation and learning, many of whose lessons still remain latent or ‘tacit’ rather than explicit knowledge. We seek to identify what factors – institutional, cultural, social and physical – make for successful outcomes in inter-city competition for that most valuable global resource of the enterprising creative immigrant, or re-inspired native, who will contribute to building a great city and by their presence make it an even more attractive place for others.

The many existing city indexes can be grouped into two classes (see Table 2 below):

The first class of index, exemplified by the work of Richard Florida and his colleagues, is based on the premise that a ‘creative class’ of migrants is drawn to cities by cultural attractors and by societies that value diversity, openness and tolerance. These indexes also strive to measure the vibrancy of the creative sectors in terms of output, employment, participation and talent. This is a stocks approach, even though the point is to attract mobile inflowing stocks of talent and intellectual capital.

Indexes in the second class, exemplified by the Global Power Cities Index, tend to include comparable (though less detailed) ‘creative’ indicators as a subset, while expanding to cover a wider pool of city attractors, including business activity, liveability, the environment, transportation and accessibility, and technology. This wider scope tends to shift the focal point from culture and creativity towards city infrastructure and basic services, innovation and technology performance, and international exchange and network formation. This is a flows approach since it measures a city by the magnitude of connections that flow between cities.

In the following review we have divided the extant city indexes into these two classes: creative stocks and global flows. We also include a further section discussing other prominent approaches from the literature on world creative cities. Tables 3 and 4 (below) presents a checklist of the dimensions and indicators covered by the various city indexes included in our review. The checklist consists of 16 general themes from which the index dimensions are drawn:

Based on our assessment of these 16 dimensions, which we take as key analytic points of decomposition:

We seek to identify the strengths and weaknesses of the respective indexes, and present this discussion in order to outline the parts of the extant indexes that we will seek to incorporate into our CCI Creative City Index, as well as highlighting gaps in the extant measures that we will seek to fill.

The following sets of index indicator proposals are less focused on creativity, as above, but more on the information economy, enterprise and trade, and technology aspects of global cities. In many cases these align with the above index measures, but the main difference is that where the above indexes ostensibly focus on the cultural life and creative potential of a city, these indexes unabashedly focus on its economic, entrepreneurial, technology and business environment, with a culture as an aspect of residency, not so much a factor of production. Our design on a good solution for a creative cities index is to take the best of the above suites, with some further inclusions and adaptations, and then to incorporate them into an edited version of the best of the more general suites below. Of these, we suggest that the Global Power Cities Index provides the best platform for development.

Substantial sets of indexes have already been created to map creative cities and global or world cities. Most of these have been developed within the past decade or so and, notably, none goes back through many decades. Global city index construction is a new, emergent industry. If we are to be guided by the evolutionary experience of industrial developmental logic, then we may reasonably expect that the future path will be a typical explosion of variety (as is already evident) followed by a competitive shakeout to arrive at a handful of institutionalised competing indexes (as appears imminent). Not all of these indexes will survive. Some may merge; others may vanish. Furthermore, some of the underlying methods and focuses may become more prominent while others are selected against, whether by a lack of demand, or by a withering of supply. It is unclear at present which of the 23 indexes we surveyed above will still be around in say 5 years. But the importance of this measure guarantees that some will still be around, in some atavistic form, in 20 or even 100 years. Now is the time in which such institutionalised definitions are being refined and selected. Our goal here, then, is to intersect this point with a proposal for a new integrated and superior index, not just another index. We believe that the opportunity and brief issued by the Beijing Research Centre for Science of Science arrives at a crucial junction in the evolutionary history of global city index construction by its willingness to address the seeming X-factor that accrues in many of the above indexes in relation to creative city dimensions.

It is not for us to say which of the above indexes might prosper and which might fall away. Some are more comprehensive and robust scientifically than others. Each has strengths and weaknesses and they are differentially useful and applicable, depending on the purpose. Instead of picking winners, we seek to integrate the lessons offered by each and any of the indexes to date into the construction of a better index.

Unfortunately, deciding what lessons are most important is not an easy matter. Few and perhaps none of the extant indexes have been appropriately tested against objective measures of actual city development and change. They are all relative indexes (‘more’ or ‘less’ measures, not measures of absolute values), which imply a projection (if City A is number one and City B is ranked, say, number 5, then A should perform better than B). But even such relativities have not been systematically and scientifically tested, not least because the newness of the indexes has not allowed for data about longitudinal trends to be gathered. At the moment the assessment of an index is not based on what is objectively accurate, as with, for example, a consumer price index that measures inflation, which will soon enough be proven right or wrong. Currently the evaluation of respective indexes is mostly about media, networks of influence, and citations, yielding relative values (City A has more of Factor K than City B). However, relativity is not uncommon in science – for instance geology was founded on the relative ages of rock, until radiogenic dating enabled absolute age to be calculated; and still relative age is important – it’s important to know if a rock is Cambrian or Cretaceous if you’re looking for fossils, or for calculating the age of inaccessible rocks (e.g. from other planets). Thus, a comparative approach using relative values is appropriate for the kind of data presently to hand.

What is a creative city? What are its aspects and properties, and how does it differ, if at all, from a global city? Is, thereby, a creative city index (Y) simply an addition to extant global city indexes (X), as in all such inclusive X measures plus this additional Y suite of measures (i.e. X + Y)? Or is that new Y suite that we propose actually an alternative formulation to X (Y not X)? Or is it really a new and complex X-Y hybrid that is proper and true? Is a creative city a different concept and measure from a global city?

Our view is that a creative city is not the same as a global city. Rather, the relation is that between a germinal seed and the resultant organism; a creative city comes first. A creative city can (but may not) become a global city. Attempts to create a global city cannot succeed unless it passes through the stage of becoming a creative city first, in some way.

We hypothesise that to become a global city, it is necessary to first be a creative city. Such a theory implies both a developmental and an evolutionary logic. We believe it to be true regardless of the initial size of the city. We suggest the historical evidence supports or at least does not reject this proposition. Every global city can be identified with a prior (and often continuous) creative apogee.

A global city is the ‘adult’ form of a creative city, in the developmental metaphor. But equally, pursuing the metaphor, one cannot create a global city ab nihilo, from bare land and imagination. The difficulty of achieving creative, never mind world, status for planned cities is demonstrated in the relation between federal capitals and world cities in, for instance, the USA (Washington DC), Australia (Canberra), Brazil (Brasilia), Nigeria (Abuja) and Burma (Naypyidaw). In no case do these forgettable capitals outshine the organically nurtured world city of their respective countries – New York, Sydney, Rio, Lagos or Rangoon. Such ventures are nevertheless plentiful, from the proposed Charter City movement (á la Paul Romer), the new ‘aerotropolises’ such as New Songdo City in Korea, to the many new cities currently underway in China.

It is instructive to recall previous doomed ventures such as the Multifunction Polis in Australia: http://en.wikipedia.org/wiki/Multifunction_Plis; and on a smaller scale the mixed career of planned 'garden cities' and 'new towns' in the UK, such as Welwyn Garden City (England), Cwmbran (Wales), and Cumbernauld (Scotland).

This applies even where the designer of the green-field plan is creative to a fault, as was the case with architect Walter Burley Griffin for Canberra. He believed that: ‘I have planned a city that is not like any other in the world. I have planned it not in a way that I expected any government authorities in the world would accept. I have planned an ideal city – a city that meets my ideal of the city of the future’ (Wikipedia: Walter Burley Griffin). And Canberra is well endowed with cultural institutions and notable architecture. But after a century it has never acted as an attractant; its creativity is planned too, not self-organised by a heterogeneous population pursuing their own entrepreneurial and creative interests. Australia’s creative and world cities remain Sydney and Melbourne; Canberra retains the atmosphere of a town for bureaucrats.

Thus we advise caution in pursuing a ‘real-estate’ model of city development. It misunderstands the forces that shape city development and evolution in the specific dimension of creativity. Cities develop organically by continual infusion of new talent and ideas, and by the annealing and selection that produces. Great cities are evolved, not designed.

Only small cities have great leaders; great cities find their own path. Great cities are the outcomes of clashes and conflict of systems of value and meaning as much as of harmony and community within such systems of value and meaning. These transplant only with great difficulty and only under exceptional circumstances. Hong Kong and Singapore became global cities through the synergistic creative energy of expat immigrants (from Britain, the Mainland, and across SE and South Asia) and local citizens (themselves of different origins) faced with open spaces of opportunity. In both cases the cities were built up rapidly from small beginnings in pursuit of trade and empire, but their dynamism and innovation came from individuals seeking to take advantage of these opportunities, not from being designed in advance. This highlights the basic limitation of the ‘real-estate’ approach: it has no built-in system of value because such systems are social and cultural complexes that are contained in cultural and institutional forms, not in buildings or street-plans. No great city can form without first being a dynamic and creative city, and those values arise from the self-organised interactions of myriad people, not from blueprints. We doubt that there are exceptions to this rule.

Great cities develop from within, but they do so by infusion from without. This is the seeming paradox of great city development. It grows from internal feedback, but it does so by infusion from external energy. The resolution of this paradox is to recognise that a great city is a complex dissipative system, as Jane Jacobs first described in 1970. What matters is access to its flow of energy in order to participate, which it to say, to be a citizen in the entrepreneurial sense – to have a voice, or at least the prospect of such. In a creative city, this is an expectation: an incomer can enter, find a voice and an audience, and may make a difference, regardless of their identity or the content of their idea.

Thus, traditional objective measures of population or market growth are a necessary but not sufficient condition for assessing creative cities. To predict a creative city, we need also to know how it facilitates expression of new ideas and how well it can adapt to forms of meaning and value that point to the future not the past. How does a city perform as an experimental space for those of its own citizens whose lives centre about such experimentation? As a demographic cohort, these are both the young, who are seeking to start their productive lives and who have weak ties to existing arrangements, and also the entrepreneurs and immigrants who seek to reinvent or improve their lives. All succeed only to the extent that they can also change the lives of those around them, which they do so only by altering their assessments of meaning and preferences over value. In a creative city, citizens must be open to change their values and preferences, to recast their assessments of meaning, in consequence of the examples and bidding of socially networked and trusted others. To the extent that this social learning process can happen, cities are constitutionally creative and can thence potentially become world cities.

World cities are therefore creative in the sense that they are spaces where meaning and value are open to contestation and where social learning can occur. They will by definition be sites of turbulence and conflict, not in a warring sense but rather as spaces of experimentation, negotiation, dissonance and discovery (see Hayek 1965, Popper 1965, Rodrik 2008, Stark 2009). Such uncertainties are navigated, from the point of view of individual agency, with ‘teleological’ purposefulness (in Veblen’s sense; see Hodgson 1998: 423) that marries maximum potential advantage to the self with maximum attention – and with that, changed behaviour – from the social network (Lanham 2006).

This balance is hard to get right for individuals, never mind societies. By its very nature it cannot be planned, for a plan supposes prior knowledge of the outcome. A creative city is an experimental space. It does not know where it is going next, which makes it a politically difficult space to inhabit. It cannot have great leaders, only great custodians. But it can indeed have great citizens. Few could name the mayors of London or Paris through the 18^th or 19^th centuries, but many people know their great citizens through that time. Similarly, many people can name famous denizens of ‘Silicon Valley’ for example but they would be hard pressed to name the political leaders of that ‘city’, much less the real-estate developers of the garages, lofts and offices in which now-global corporations had their beginnings. Creative global cities are known by the achievements of their citizens.

In order to capture the dynamics of population-wide creative productivity in relation to meanings and values, brokered through the competitive clash of different institutions and networks that nevertheless together comprise the overall complex dynamic systems which compose creative cities, we need to use the insights discussed above to generate a new approach to the Creative City.

First, what a creative city is not. It is neither a medieval city nor a modern one. Both of these types maintain their attractions, especially for tourists, but in relation to creative dynamics they are both examples of ‘sunk capital’. Table 9

Adapted from Hartley 2009: 16-17; © J Hartley.

Recalling the discussion of the difference between ‘power’ cities and ‘enterprise’ cities that we identified at the outset of this report, the ‘attractant’ or creative city therefore emerges from the globalising values on the right of Table 10 below, rather from the premodern and modernising ones on the left:

In relation to the specifics of the arts and creative culture, these oppositions play out as follows (Table 11) – again, a creative city (as opposed to a powerful city) displays the values on the right:

In relation to the historical dynamics of mediated communication, globalised creative cities are digitally based, differing from previous epochs as the terms in Table 12 indicate.

In order to capture this emergent, future facing or ‘attractant’ city, we focus on the growing importance of youth. As a cohort, they are willing agents of risk and uncertainty, operating as both producers and consumers of experimentation and novelty in a social-network system of cultural entrepreneurship and consumer creativity. Further, as a global demographic, youth is under-recognised in much contemporary policy work, which tends to assume it is catering for an older demographic, from which elite arts producers and consumers are drawn. There is also much written about the aging demographic in many countries, including China, and certainly the overall profile of humanity is aging. However, entrants to the future workforce, those who are most interested in experimenting with novelty and willing to risk change, and who are currently educating for a creative, enterprising, and increasingly urban future, are predominantly drawn from the 0-24 age group. By 2050 there will be around 3.5 billion people under 25 worldwide. Of these, almost 1.65 billion young people will live in China, Europe, India and the USA alone, and another 1.8 billion – a majority – in all other countries:

Figure 4

Among those youngsters, the current estimate of people between the ages of 10 and 25 – the group most active in future forming – is more than 1.8 billion. This is what the UN calls a ‘demographic dividend’, providing the ‘labour and skills needed to rebuild cities, economies and lives’ (UNFPA 2010: 47-9). One may add that they will also be the cohort that reinvents the city, through their own ideas, usages, and interactions.

While providing for young people is a challenge, an aging population is also a challenge, particularly for innovation dynamics. Here, counter-intuitively, China may have more to worry about than the USA. According to one observer: ‘Surprisingly, the process of population aging is much less dramatic in the United States of America than in China or Europe – both due to higher fertility and a large number of (young) immigrants’ (China-Profile).

See: www.china-profile.com/data/ani_ceu_pop.htm.

We believe that dynamic, future-facing creative cities will be characterised by a relatively high total number of both residents and immigrants in the youth cohort, and that there will be increasing pressure on China to compete internationally. Furthermore, by one calculation, Beijing currently ranks only 20^th among the world’s ‘agglomerations’ of total population (all age-groups), after Tokyo (34.3m), Guangzhou, Seoul, Delhi, Mumbai, Mexico city, New York (incl. Newark, Paterson), Sao Paulo, Manila, Jakarta, Shanghai, Los Angeles, Karachi, Osaka, Kolkata, Cairo, Buenos Aires, Moscow, Dhaka and Beijing (14m).

Source: www.citypopulation.de/world/Agglomerations.html. By contrast, London ranks 24th; Paris 26th; sydney 75th.See also:www.citypopulation.de/cities.html.

The CCI Creative City Index proposes indicators for the components of a city’s population (both existing residents and incoming immigrants) that may yield a ‘demographic dividend’ for creativity while recognising that mere size isn’t everything. This requires a focus on the 10-25 years old cohort (teens and young adult), with indicators clustered around the main ‘attractant’ affordances for this cohort. Such attractants will certainly include aspects of ‘youth culture’ – media, music, festivals, clubs, crowds, scenes, malls – but these may not be the most important resources. Philosopher René Descartes dubbed Amsterdam, an early ‘world city’, an ‘inventory of the possible’ (Brook, 2008: 8).

T Brook (2008) Vermer's Hat: The Seventeenth Century and the Dawn of the Global World. London: Profile Books.

A recent article by the Financial Times’ architecture critic, Edwin Heathcote, has made the case for what attracts people to cities – it isn’t a ‘liveability’ index (routinely won by ‘boring’ cities) but ‘complexity, friction and buzz’. He prefers ‘places which are large and complex, where you don’t know everyone and you don’t always know what’s going to happen next; places of opportunity but also of conflict, but where you can find safety in a crowd.’ He asks, can poor people and immigrants ‘improve themselves, reinvent themselves? Is there upward mobility?’ He cites London and New York as ‘magnets for immigrants precisely because they allow those kinds of new beginnings.’

E Heathcote (2011) Liveable v lovable. Financial Times. www.ft.om/cms/s/2dd9bba18-769c-11e0-bd5d-00144feabdco.html#ixzz1LpVGLjVB. In the ensuing discussion of 'which is the best city' Beijing was not mentioned (nor any city in Australia).

The following indicators (Table 13) are carried over from the index library surveyed in Chapter 2 above. Many of these ‘themes’ are incorporated and merged into our final index construction.

The following sections 3.5 and 3.6 outlines the indicators that we used as a starting set from which we then refined our eventual indicator suite. These are based on the indicators that we derive directly from the CCI model as well as a suite of ‘additional’ indicators that we derive from the other indexes reviewed above.

We present this full set here because many of the eventual refinements in the six-city survey were partially based on data availability, data consistency or cost considerations, and a more expansive or comprehensive index would seek to re-include these considerations.

Because these classification sets and indicator suites are based upon our theory of the creative economy it will be instructive to summarise here what we mean by a creative economy as derived from our focus over the past decade at the CCI on the creative industries. In brief, we have identified four distinct versions of the creative industries as associated with different classes of definition. They are:

CI-1: Creative Clusters

CI-2: Creative Services

CI-3: Creative Citizens

CI-4: Creative Cities

We are proposing that the four phases noted above, CI-1 to CI-4, also form four distinct models, not based on trying to define ever-more tightly how creativity is an industry but, on the contrary, on showing how creativity needs to be accounted for at ever-increasing distance from industry. It is not until we reach stages CI-3 and CI-4, where creativity reaches cultural dimensions located in cities, rather than being confined to production processes located in firms, that the connections between culture and economy, individual talent and societal scale, can come into focus.

Furthermore, it is only at that point that we can take proper account of the growth of ICTs, digital media and the Internet, because these are now not simply in-company efficiency-technologies (as IT once was), but whole-of-society cultural forms (as the Internet now is). In other words, if we confine the notion of creative industries to the traditional (i.e. analogue) creative arts and their industrial or occupational form, we cannot account for the importance – both economic and cultural – of user-created content and the burgeoning scale of computer-enabled social networks. Since these are clearly important drivers of the creative industries, we need all four models before we can develop an index to explain creative innovation, never mind the integration of cultural and economic meanings and values.

Finally, CI-4 reminds us that these developments are competitive and uneven across space as well as time – some cities are creative innovation ‘spikes’ compared with others. Some important examples and points relating to the interaction of these four levels in a creative city are summarised below.

Interaction of CI-1 and CI-2 with CI-3 and CI-4

Based on the CCI Creative Industries Model and the extant creative cities literature hitherto reviewed, we can highlight numerous other issues and potential indicators for consideration. Tables 19-27 outline these aspects in terms of:

Infrastructure and Institutions

Creative cities have the infrastructure, institutions and public capital for the production and development of new ideas.

Interaction of Agents

Creative cities link consumers and producers. This means a vibrant local scene of consumption as well as production.

Cultural Complexity, Clustering and Networking

Creative cities build the social connections and networks along which social learning and copying enables creativity to spread, replicate, and develop.

Creative Milieu

Creative cities bring together complementary practices of cultural production and consumption.

‘Virtual’ Clustering and Networks

Creative cities are digitally connected. Citizens have digital literacy.

Population Characteristics

Creative cities have populations that tend to be energetic, young, ambitious. These can be characterised by early adopters open to experimentation and novelty.

Formal and Informal ‘Creative’ Education

A creative city develops formal and informal education pathways. Formal measures would include credentialed ‘fine’ arts education or film or dance schools for example; informal creative education would include participation or media exposure and experience.

Clash of Different Systems, Generations, and Cultures

A creative city will tend to be a complex space that generates interactions of ideas from different demographics, cultures and economic activities; creativity emerges from the clash and conflict of differences. It will have attractive natural and built environments (and the criteria of ‘attractiveness’ will differ for different demographics), a diverse range of people, and vibrant street life (see cover picture).

Urban Environment and Attractions

A creative city will have a level of amenities and a sufficiently attractive and interesting urban environment that induce people to move to the city because of the lifestyle.

The CCI Creative City Index comprises 8 categories of sub-indexes:

We contend that these eight categories, which will be discussed below in relation to their pilot estimates, define dimensions that are both sufficiently uncorrelated yet additively complementary aspects of what makes a ‘creative city’.

This chapter outlines the findings of the estimates of the CCI Creative City Index and offers some discussion and justification for the various indicators and measures involved.

At present, the index has been tested on six cities (Brisbane & Melbourne in Australia, Berlin & Bremen in Germany, and London & Cardiff in the UK). These cities were chosen to provide a mix of countries with differing economic and industrial backgrounds. We also chose one large and culturally dominant city (London, Berlin and Melbourne) and a second tier city that was nevertheless still a significant economic and cultural hub (Cardiff, Bremen, Brisbane).

While the pilot would have been strengthened with additional city estimates for developing country cities, such as in Indonesia or Brazil for example, budget constraints and language limitations mean that we have only tested the results on these six. There remains much further work to develop and extend this index to other cities and countries.

The unweighted (or equally weighted) index results over the eight index dimensions are presented in Table 28 below.

Indicator Definitions and Theory

The first sub-index is a familiar measurement of creative industries and services in terms of scale, scope and employment. There are three sub-levels to this category:

o Creative Industries Scale (CI-1: Creative Clusters)

o Creative Industries Scope (CI-1: Creative Clusters)

o Creative Industries Employment (CI-1: Creative Clusters; CI-2: Creative Services)

Creative Industries Scale comprises two measures: the size of the sector in economic terms (GDP) and the number of firms in the sector. These are indicators of Creative Clusters (CI–1 in the CCI creative cities model). The CI scale measure and also the employment measure below are both constructed in per capita terms.

The Creative Industries Scale classification is made over six categories (see Table 32 below, available at http://eprints.qut.edu.au/8242/1/8242.pdf), which reflects the CCI use of a pragmatic approach to identify the segments containing occupations and activities that are either closely related to each other or closely aligned in a production chain.

Creative Industries Scope as a measure of the creative city is unique among existing indexes. No other creative city indexes attempt to include this in their construction. We suggest that along with size, more creative cities are characterized by a more diverse creative sector in the sense of a greater balance of all creative sectors and tend not to be dominated by one particular sector (e.g. film-making).

Creative Industries scope focuses on diversity within the sector because of a long-run connection between diversity and growth. A creative global city is presumed to have a diverse industrial ecology in creative industries and we have sought to estimate that in our index construction. This alludes to the specialization vs. diversity debate in industrial and urban economics and we claim here, following the findings in that broader field, that diversity trumps specialization. The work of Harvard urban economist Edward Glaeser, among others, have found that over decades, industrial diversity contributes more to economic growth than specialization. Although we are not necessarily concerned with the impact of creative industry diversity/specialization on economic growth, we are making the parallel argument that, over decades, creative diversity contributes more to creativity growth than specialization.

We use standard population diversity measurements. The Shannon index has been a popular diversity index in the ecological literature and estimates the proportional abundance of sub-sectors in the overall population of all creative industry firms. The Shannon index also measures the uncertainty involved in cultural clash, contestation and mixing: the more different sub-sectors or firms there are, and the more equal their proportional abundances in the greater creative industry of the city, the more difficult it is to correctly predict which firms or sub-industries will be encountered. The fragmentation index is another common population diversity measure that we have adapted here (with much the same interpretation). We take the average of these two diversity indexes as the measure of scope. While these are perhaps somewhat abstract index measures, we suggest that such diversity indexes capture both the industrial resilience of a creative city (a high diversity index associated with greater resilience through greater potential for growth along in any part of the sector) and also for greater potential for recombinant growth through new combinations of different parts of different sectors.

Creative Industries Employment comprises just the one measure: the number of people employed in the sector, which is then rescaled to a per-capita measure. This is an indicator of Creative Services (CI–2 in the CCI creative cities model). Note that we use per capita measures not absolute measures because we do not want to conflate general non-linear effects common to all cities (e.g. the benefits of larger population in cities such as London and Melbourne marked by higher aggregate CI output and employment) with the true local dynamics specific to each city, for example the effects of local events, historical contingency, or ‘creative intensity’ that are independent of population size.

Methodology and Data

The data can generally be found on the national level and are by now standardised. We transformed the data for Germany and the UK to fit the Australian standard of six categories. The data are generally available on the national level but not on the city level. However, the statistical authorities in Australia, Germany and the UK do also collect these data on an urban scale and they should be made public in the near future. For instance, the German Ministry of Economics and Technology will publish these data for the bigger cities from June 2012 onwards. In order to estimate the size of the creative industries in one city we scaled the state data (or national in the case of Bremen) down to city size by using a conversion rate. In this case we used the OECD data on patents to establish the relationship ‘patents from city/patents from entire state (or nation)’. This coefficient was then multiplied with the state data to get city-level data. This method was then applied consistently where we needed to transform national data to city data.

Scope is calculated with a Shannon’s Diversity Index measure. This is a standard method for calculating a diversity index from a simple unbounded distribution. A higher result implies a more equally distributed share of each industry out of the total. Notably, this measure does not account for absolute scale differences across different cities, but only relative differences between sizes within a city. This index therefore works best when comparing cities of similar sizes. To obtain the diversity index measure the percentage of each industry out of the total creative industries is calculated. This share is then multiplied with the natural logarithm of itself. Finally, the sum of these numbers for every individual industry is then multiplied by negative one. The following formulademonstrates how a Shannon’s Diversity Index is calculated:

D i v e r s i t y = −<!-- − --> ∑<!-- ∑ --> i = 1 N p i log ⁡<!-- ⁡ --> p i $${\it { Diversity }}=-\sum_{i=1}^{N} p_{i} \log p_{i}$$

Results and Discussion

The results are shown in the table below (Table 33). The analysis comes as no surprise – London has the biggest creative industries in all aspects. The following ranks are the same as the population rank, i.e. Melbourne comes second, Berlin third, etc. When it comes to the scope of the industries the picture looks different. Bremen and Berlin are the most diversified cities but the difference from other cities is not big. The structure however is not the same. Melbourne’s number one industry is software followed by the publishing and print media. Berlin’s music and performing arts industry is nearly four times that of Melbourne. Although Melbourne’s creative industry is bigger, Berlin’s is more diverse.

Indicator Definitions and Theory

The second sub-index is another unique contribution of the CCI Creative City Index, ‘microproductivity’, which measures population-wide microproduction by creative citizens, virtual connectivity between agents, and local networks and interaction. There are three sub-levels to this category:

Population-wide Microproduction as a measure of the creative city is unique among existing indexes. No other creative city indexes attempt to include this in their construction (that we are aware of). This indicator of Creative Citizens (CI-3) is a key component of the CCI creative cities model, and we use various measures to capture it (as below). These metrics attempt to account for the importance – both economic and cultural – of user-created content and the burgeoning scale of computer-enabled social networks. We prefer to use actual production data (e.g. the amount of citizens’ uploads to YouTube in London); however in the absence of such information we have used ‘tagging’ data from scaled-up micro-productive institutions (e.g. YouTube, Google) as an approximation.

Indicators we have used include:

A feature of microproductivity is that it is network-based. The next two sub-components measure microproductive networks – in virtual and real, physical spaces. When creativity reaches cultural dimensions located in cities, rather than being confined to production processes located in firms, the connections between culture and economy, individual talent and societal scale come into sharper focus. It is at this point that we must now take proper account of the growth of ICTs, digital media and the Internet, because these are now not simply in-company efficiency-technologies (as IT once was), but whole-of-society cultural forms (as the Internet now is). Virtual Connectivity comprises numerous measures to capture this aspect of the CCI creative cities model (CI-4: Creative Cities). It is about connectivity infrastructure or capability but also about virtual sites for social meeting and mixture, as well as friction.

Indicators we have used include:

If Virtual Connectivity is the digital face of the creative city, then Local Networks and Interaction is its analogue or real-world counterpart (CI–4: Creative Cities in the CCI creative cities framework). It is about the local places and events that facilitate the associative structures and social networks, connections and human interactions that characterize creative cities e.g. physical sites for meeting, mixture, and friction.

Indicators we have used include:

Methodology and Data

The data for sections 2.1 and 2.2 are fairly easy to obtain through simple Internet research. It is important to take comparable statistics, i.e. British Facebook statistics are best of all compared to Renren Wang statistics in China. For this index we worked with ‘tags’ as they are easy to get and to compare. There is the risk that popular cities are tagged more often (e.g. by visitors) than less popular cities. On the other hand, it could mean that ‘attractive’ cities (as discussed in 4.3) generate more creativity. Better data could be obtained if the providers of Internet services could be contacted directly to supply actual uploading data from a given city and not merely tags. This is however a matter of dealing with the companies, which can take a considerable amount of time. Data about the accessibility to the Internet and the type of connection are usually published by statistical bureaus or local trade ministries.

We wanted to collect data on participation in associations (which may include democratic, cultural or health & welfare purposes), but Section 2.3 is possibly the hardest section to measure, as this kind of information is very different in every country. German cities for instance measure the number of people who are involved in some form of sporting association, but this information is not published in Australia. On the other hand, the Australian Bureau of Statistics reports on voluntary involvement and the number and amount of tax-deductible donations made by Australians, but one would search in vain for a German publication on this topic. Another issue is the nonavailability of certain services in different countries. Meetup.com for instance is widely used in Anglo-American countries and all kinds of interest groups can organise activities. In Germany this website is not widely used and there is no counterpart that could be substituted for comparison. Another good measurement is the number of members of the local Chambers of Commerce; these however do not disclose such information. The ideal variables for section 2.3 would be the number of associations (split up according to different fields, i.e. sports, culture, business, charity, etc.) and their members as well as the number of activities organised by them (activities, fairs, tournaments, etc.).

Results and Discussion

London has the highest score, followed by Berlin and Melbourne. It is often tagged which has to do with the attractiveness of the city. Since many things happen in London it gets referenced a lot, which brings the ‘tag’ score upwards. Berlin being the German capital has the advantage of being often quoted while Melbourne, which is bigger than Berlin, does not have the capital-city benefit. Berlin is nonetheless more creative than Melbourne, which is remarkable for a city that has only three million people in comparison to nearly five million in Melbourne.

The virtual connectivity figures demonstrate a difficulty with cross-cultural comparisons. Germany has traditionally had several private and professional networking platforms. Therefore, the German market has been more fragmentised as its British and Australian counterpart. This biases the results downwards. The simple sum of different services cannot be taken since many users use several platforms and a mere aggregation would lead to double counting and overestimation.

Berlin has an important music scene and a vast number of associations that engage in the organisation of festivals. This explains why there are more activities there than in Melbourne, which has a significantly bigger population. Other cities like Brisbane, Berlin and Cardiff are dwarfed in comparison to these big cities, but it is not surprising that scale has its effect on the result.

Indicator Definitions and Theory

The third sub-index comprises familiar measurements of creative infrastructure and attractions. However it also includes indicators relating to the ‘economy of attention’ (Lanham 2006) that surrounds a city. There are two sub-levels to this category:

Creative Attractions contains various indicators, most of which are straightforward. These range from tourist attractions and ‘things to do’ to the number of concert halls and radio stations. We also highlight the importance of great cities as ‘zones of attraction’. In this sense causes of city growth are the outcome of the choices of a mobile elite of smart global citizens. The issue then becomes how to attract such great people. This of course means that cities compete, and Economy of Attention indicators are one way of measuring which cities have a more visible ‘buzz’ to them and advantageous cultural pull factors.

Unlike scale and scope measures in dimension 1 above, which were transformed to a per capita index to eliminate pure scale effects, creative attractions and economy of attention is not transformed in that way but remains as an absolute measure of a city’s attractions and the attention that it receives. Here we are explicitly interested in the full-scaled appeal of the city: cities compete for internationally mobile, creative citizens, and they compete of size, on the aggregate picture of attractiveness.

Economy of attention is quite similar to the “city name tagging” measurements in Microproductivity, however it is different in the sense that these indicators are more about promotion of the city and often originate from more “traditional”, professional creative industry sources (rather than microproduction from the network, pro-production from firms). These indicators are thus about the likelihood of a person encountering a city in popular culture; the general buzz around a city, and how likely you are to hear about or to know something about the city.

Indicators we have used include:

Methodology and Data

As can be seen, the vast majority of variables can be searched on the respective websites. The number of hotel rooms and cinemas, if not available from the statistical authorities, can be easily researched through specialised websites. The city council websites offer comprehensive lists of art theatres, museums, concert halls and libraries. Libraries can otherwise also be searched for in the statistics of the national library

associations. CB Richard Ellis Global Research publishes the number of top retailers in a city.

Results and Discussion

London attracts most attention, followed by Berlin, Melbourne, Brisbane, Bremen and Cardiff. Melbourne has a bigger infrastructure but Berlin manages to receive more attention in publishing, almost certainly because of its record as a pivotal point in European history. It seems that more tourists come to the fairly central European city, as it is more accessible than the city at the southern tip of the Australian continent, thus more hotel rooms are available in Berlin. It is noteworthy that Berlin has most cinemas of all but that their size is small. Small cinemas are more likely to show independent movies than big cinemas that can only fill their seats with blockbusters. Small cinemas are also likelier to show local productions. The same principle seems to count for the number of live venues, concert halls and performing art theatres.

Indicator Definitions and Theory

Cultural Participation is perhaps the most straightforward sub-index. It contains two sub-levels:

These are about support for the creative industries and culture from citizens (as opposed to the government) in the form of attendance (patronage, involvement) and expenditure. The number of admissions to selected cultural events measures Attendance and annual household expenditure on arts and culture measures Expenditure.

Methodology and Data

Participation and expenditure data are as hard to find as ‘microproductivity’ data. Many of the following numbers are simply not recorded. The authorities of Berlin, London and the Australian states offer reasonably useful numbers on participation, but Bremen and Cardiff and the individual Australian cities publish barely anything on this. Alternative sources of information are the relevant venues themselves although they often do not record these statistics either and for bigger cities this becomes a very tedious task. Household expenditure data can usually be found in statistical yearbooks for countries and eventually states but does usually not exist for smaller cities as it is hard to measure and very few people are willing to participate in these kinds of surveys. In this case only the state or national level data can serve as an approximation.

Results and Discussion

Most participants in cultural events can be found in London, followed by Berlin and Melbourne. The difference between Berlin and Melbourne becomes particularly apparent here. The participation at ‘Total GLAM’ is effectively biased against Berlin, because only the numbers attending museums could be found, while all kinds of attendance (i.e. GLA as well as M) could be found for London and Melbourne. If the attendance at libraries, galleries and archives was added for Berlin, the difference would be even greater. Looking at the share of the expenditure on arts out of total household expenditure the picture becomes clear. Cultural activities in Berlin are cheap in comparison to other cities. Furthermore the living expenses in general are low, so a bigger proportion of income may be dedicated to culture and arts. The same is true for Bremen, where many participation statistics could not be found and the city is thus undervalued.

Indicator Definitions and Theory

This sub-index measures public support at each level of government in terms of:

Again, this section contains very straightforward measurement of the amount of cultural funding per person by level of government. We might also like to look at the proportions of the local city government budgets that are devoted to creativity and cultural funding and ICT funding as an alternative indicator.

Methodology and Data

Government budget reports at national, state and local level are the obvious source for this measure. Unfortunately it can be challenging properly to disentangle all statements and to properly attribute expenditures to either mere bureaucracy payments or real investment in culture. The budget of the ministry of culture is generally a good approximation. This section would also include ‘government expenditure on creative industries’ if it had proven possible to attribute the payments to these industries.

An alternative measure of public support could be the proportion of local city government expenditure on arts, culture and creative industries (spend on culture, arts and creative industries/total spend). This would measure the relative importance place on cultural expenditure in government budget. This was difficult to quantify across the various levels of government; future effort could be put on refining this as an indicator.

Results and Discussion

The range of spending on culture and arts per capita is rather narrow, with Brisbane and London spending the most. Cardiff is unfortunately biased downwards because the local government’s expenditure could not be found.

Indicator Definitions and Theory

The next sub-index is another reasonably familiar measurement of R&D employment, education and human capital, and research capacity. There are three sub-levels to this category:

The first indicator is simply the number of people employed in R&D. Education indicators are common to the city index literature; ours includes the number of students, graduates and education institutions. Finally, research is measured in the standard way used in other indexes by the number of patents issued and R&D expenditure.

A key part of our conception of a creative city is youth and population dynamics – the growing importance of youth. There is an obvious overlap between the youth/student and microproductive demographics: students are often willing agents of risk and uncertainty, operating as both producers and consumers of experimentation and novelty in social-networks of cultural entrepreneurship and consumer creativity. This is the very definition of microproducers.

Moreover, as a manifestation or symptom of the gap between “elite” arts producers and consumers and “microproducers/consumer-creators”, youth is often under-recognised in much contemporary policy work, which tends to cater for an older demographic. By measuring the number of higher education students as a promoter of the creative city we hope to restoring some balance to creative city policy landscape: from elite producers (older demographic) to microproducers (student demographic). Besides, it is also important to have a vibrant student population from a purely educational and human capital perspective.

Methodology and Data

National and state level statistics are generally a good and reliable source concerning education (6.2). Occasionally some of the big panel surveys from social research projects can help find statistics for individual cities. Section 6.1 on employment and section 6.3 on research and development are best of all taken from the Organisation for Economic Co-operation and Development (OECD).

Results and Discussion

Since all six cities are in industrialised countries the difference between the cities is not as big as in other sub-indexes. London has the highest index figure as it is home to many corporate and public research centres. Also the vast number of students and the facilities give it an advantage of scale. Berlin is again ahead of Melbourne as it can compensate the disadvantage in scale with better relative factors. It is interesting to notice that Berlin outperforms London in ‘number of patents’ and its per capita value. Bremen is slightly ahead of Brisbane and Cardiff comes last, despite its sizeable student population. It is interesting to notice that the German cities have the lowest number of university graduates out of the total population ratio while they have the highest proportion of qualified people. For the UK the opposite is the case. This could have to do with different alternative training programs like apprenticeships, which are fairly well established in Germany and Australia.

Indicator Definitions and Theory

This sub-index measures global integration in terms of:

We have highlighted the shift towards the modern globalised and entirely dynamic view of great cities as zones of attraction. This view emphasizes the role of cities as international hubs, or nodes that mediate local networks with global networks and connectedness. There is a focus on growth focused transport networks, international air connectivity and the cultivation of ‘aerotropolis’ cities. Flows of People is measured by net overseas and interstate migration, and population turnover by migration (total flows/population). Globalization is measured as well – the most well known of these indicators is the Globalization and World Cities (GaWC) group ranking.

Clearly international airport traffic is a key part of global integration. Indicators we have used include:

The sub-component measuring flows of people based on the premise that cities compete for ‘factors of production’ i.e. in effect, people and capital. Creative cities compete for the creative class and we have outlined some attractions and the economy of attention involved in this process; however, another point is that the creative city itself is an attractant of creative people (indeed that is the point of the CCI-CCI). We want to measure the extent to which this population turnover and flux occurs not only as an output of the creative city (effect) but also as an input (cause). That is, we make the claim that people are attracted to places in flux.

Indicators we have used include:

Globalization is measured by two indicators from the GaWC group. GaWc’s research network focuses upon the external relations of world cities as opposed to simply studying the internal structures of individual cities and comparative analyses of the same. These indicators are based on the network of business relations between the world’s largest law, financial services, advertising and accountancy firms. GaWc globalization indicator is based on the ranking and classification of world cities given by GaWC whereas the number of ✓connectives" is the actual number of connections between firms in the cities.

Methodology and Data

International airport traffic statistics are available from virtually every airport authority, whose statistics are usually quite comprehensive. Most state-level statistical yearbooks have a section on interstate and overseas migration, although these could be reduced to only net values. To deduce city data the in- and outflows show be scaled by population size. The GaWC Globalization Index has 12 categories of cities. Depending on which category the city is in it receives this score, which then merely needs to be scaled into a percentage.

Results and Discussion

This subcategory mostly depends on aggregate statistics. Thus it is not surprising that bigger cities will have a higher score. London again leads, followed by Melbourne, Berlin, Brisbane, Bremen and Cardiff. Berlin does better than Melbourne when it comes to international flights and passengers, but this is because Germany is much smaller than Australia and has more direct neighbours. Cardiff has the highest population turnover due to its big student population who come for studies and leave again after graduation.

Indicator Definitions and Theory

The final sub-index – Openness, Tolerance & Diversity – is very common to the creative cities index literature, as popularized by the work of Richard Florida. There are three sub-levels to this category:

The indicators below can be used to measure how open a society is, with respect to either outsiders, or dissenting opinions and practices. A more open society is believed to be more conducive to creativity. The standard indicator of openness, tolerance and diversity is the ‘gay index’ – the number of persons in same sex couples as proportion all persons in couples, relative to national average. We have extended our analysis of diversity to include not only ‘sexual’, but also ‘spiritual’ and ‘ethnic’ diversity (i.e. using fragmentation and diversity indexes), as well as including measures of youth population dynamics and income equality. Also, within the global creative city literature civic engagement is also thought to be conducive to creativity.

Indicators we have used include:

Methodology and Data

Every statistical yearbook should contain most information needed for this index. Numbers on refugees are available with the United Nations High Commissioner for Refugees (UNHCR) or the ministry of interior and immigration. The reporters without borders Freedom of Press Index is available for the national level. Corruption figures are taken from the Corruption Perception Index by Transparency International. Voter participation is made public by the election committee. The Gini coefficient is often not published but numbers on income inequality are, from which the Gini index can be estimated. The only figure that is unreliable is the sexual openness indicator. In countries with registered same sex partnerships the figure is strongly underestimated since non-register same sex partnerships are not accounted for. Australia’s Bureau of Statistics published a different number that is even less reliable as the estimation method is wrong (e.g. two housemates of the same gender who respond that they are in a relationship are thus counted as homosexual). However, in the absence of a better measure we need to resort to this number. The diversity measure for religion and nationality is again calculated according to Shannon’s Diversity Index:

D i v e r s i t y = −<!-- − --> ∑<!-- ∑ --> i = 1 N p i log ⁡<!-- ⁡ --> p i $${\it { Diversity }}=-\sum\limits_{i=1}^{N} p_{i} \log p_{i}$$

Results and Discussion

The results are mixed here, although they may be plausible. The difference between Cardiff (60.8) and Melbourne (76.3) is not large. Looking at section 8.1, Berlin seems to be more tolerant than other cities. The high rate of people without religion is probably

due to the socialist heritage as well as the need to pay extra taxes when registered with the Christian churches. Australia and the UK have a longer history of naturalising immigrants and thus score well on the nationality openness score. Nonetheless, Cardiff only recently started to attract more foreigners as can be seen in 2.2. As a prominent student city, Cardiff has the highest proportion of young people, who tend to move away after finishing their studies and thus cause strong population turnover. German cities on the other hand suffer from the aging of society, where especially Australia looks better. In terms of income inequality Berlin performs best, which is again due to its socialist heritage and its comparably low-income level. Cardiff on the other hand has a lot of students who do not earn income and thus increase measured income inequality. The results for civic engagement are similar, which is not surprising for developed countries.

The proposal for a new type of index – which we have delivered here as the CCI Creative City Index or CCI-CCI – is based upon recognition of the strengths of existing creative and global city indexes. It seeks to maintain the useful and effective elements of those measures, while also acknowledging and seeking to correct their weaknesses.

The strengths of the extant indexes – including Richard Florida’s creative city index, the MORI index, the GaWC index, Landry’s creative city index, and others – revolve about measurement of creative city attractions, infrastructure, research and human capital, public support for arts and culture, public participation in culture, and openness, tolerance and global connectedness. We have sought to reproduce those same aspects in the CCI Creative City Index. However, there are several critical aspects of the creative economy and creative society that are commonly not included in other indexes, and it is these aspects that we have sought additionally to include (at the price of scaling back some of the above measures).

At the core of this strategy is the role of what are variously called ‘youth culture’, ‘consumer co-creation’, ‘digital literacy’ and other factors that relate to the role of the creative and engaged citizen (and not just passive consumer) in making and producing (i.e. uploading and not just downloading) creative cultural content. The significance of this point is that our own research at the CCI has underscored the role of the creative citizen, and especially those from the margins, including the young, in recreating culture and the industries and economies that are built upon it (Hartley 2009; 2012; Potts 2011). As we have emphasised, a global city must first be a creative city, and a creative city is invariably powered by energy and the entrepreneurial experimentation of the young, of the outsider, of those seeking to create new ideas and to challenge existing ideas. A creative city will invariably be complex and challenging, ‘lovable’ more than ‘liveable’, edgy rather than middle-of-the-road, often with a clash of cultures, demographics and ideas in its mix. Our index has sought to integrate these measures, including for example measures of youth and student populations as well as uploads and social networks for music.

The index has eight main dimensions, each with multiple weighted components (between 1 and 14), making for 72 distinct classes of measures. Furthermore, each of these is often composed of multiple measures (with between one to as many as ten individual distinct elements). Over 250 distinct measures are included in the index for each city. Specifically, consider the dimension and number of weighted measures:

DIMENSION NUMBER OF WEIGHTED MEASURES

This provides for a rich index that captures many aspects of a city’s creative life, economy and potential.

Our findings are limited to the data from the pilot study of six cities, so it is too early to generalise from this. With respect to the rank order and index magnitude these findings were broadly in accord with expectation. London was expected to come out well on top, as it did. Of particular interest was the extent to which London dominated, often by an order of magnitude, and sometimes more. This is a feature that was not often seen clearly on many previous city indexes, which have tended to focus on variables that have less magnitude of per capita variation in them (such as number of hotels), or tended to cluster together (such as measures of openness).

On the ‘economy of attention’ dimension, London, and then Berlin, completely dominate the other cities, including Melbourne. It may indicate that the difference between cities at the head of the scale differ from those at the tail exponentially, according to a power law scale rather than a linear one, resulting in a ‘winner takes all’ profile of leading cities compared to ‘long tail’ cities. This same ‘super-scaling’ pattern was also observed in microproductivity. These findings indicate that there is something special about cities that are both highly creative and simultaneously global cities, in that they are more than just bigger versions of smaller ‘creative cities’.

What does this imply for strategic policy? Note that the CCI Creative City Index is calculated as the unweighted average of its sub-indexes (although weightings could be applied to the sub-indexes, if required). Also note that these scores are relative to a perfect score of 100 that would be achieved by a hypothetical ‘best practice’ combination of all six cities. In this sense, the index scores can be re-interpreted similar to the distance towards the production frontier in productivity analysis. That is, the scores represent the distance towards the city creativity frontier for Brisbane (48.7), Melbourne (53.2), Berlin (63.3), Bremen (48.6), Cardiff (44.5), and London (85.1). Thus London for example could expand its Creative City Index score by about 15% by improving its performance in any of the indicators where it is not already ranked first (e.g. by raising the level of cultural and arts funding per person by level of government to that of Brisbane). On the other hand, Brisbane can more than double its CCI-CCI score and move towards global creative city status by catching up with London in the majority of the 8 categories. Indeed, this is usual the aim of high-level government funding; and this index can assist policy-makers by providing robust, efficient, evidential data to measure comparative changes over time.

While these findings in themselves are interesting and illuminating, a proper analysis would require a larger sample of cities to be studied, as well as follow-up cross-checking and statistical analysis. This remains future work. At this stage, we believe that the proof-of-concept of the index has been successfully undertaken and that the CCI Creative City Index offers a best practice platform for the benchmarking of creative city rank and progress.

One purpose of this report is to provide for our partners in the BJSS a template for the application of the CCI Creative City Index to Beijing in particular, and to other Chinese cities in turn. Many factors included in this index will be straightforward to compile, especially for the factors relating to demographic profile, industrial structure, creative industries economy and employment, public spending, cultural infrastructure, cultural attractions, education and research, and so forth. Many of these same factors also feature prominently in alternative creative city or global city indexes.

The more challenging estimates will likely accrue to the factors that centre about microproductivity and the economy of attention. Having said that, these are actually quite easy to collect, as they are amenable to web-based search. The challenge, instead, lies in finding appropriate comparisons with the proxies used. In the dimension on ‘attractions and economy of attention’, for example, we used a raft of measures that pivoted off very specific sites such as Wikipedia, IMBD (movies), Amazon (books), iTunes (music), Google Trends (Internet search), Lonely Planet (youth tourism). It makes sense for these largely American sites to be the focal points in Australia, UK and even Germany, but given the difficulty or lack of success these companies have had in operating in China it would make sense for them to be replaced with local equivalents. While this is pragmatically necessary, it also makes comparisons across a global index more difficult without adjustments for scale, penetration, and so forth.

The same issue also arises with the second dimension ‘microproductivity’, where the index we have constructed pivots off a number of nominated but dominant digital/web sites that may have no equivalents in China. For example we have measured uploading from photo sites such as Flickr and Photobucket, presence and page creation on music fan-sites such as Myspace and Bandcamp, and membership on social network sites such as Pitchfork. None of these has a strong or dominant presence in China, and to that extent they are inappropriately applied there. Instead, local equivalents will need to be constructed.

Two other domains pose potential problems. The first relates to the standard classifications and definitions of industrial sectors and creative industries. The definitions we have used in Australia, Germany and the UK are all more or less the same, and are based on satellite accounts recently used by national statistical authorities. These are influenced in turn by the 11-sector DCMS classifications (proposed in 1998) of the creative industries. China’s classifications are not always identical to this, and in part tend to weigh toward the traditional cultural industries (including cultural tourism) and to include less of the more commercial and digital industries (e.g. software and video games in particular).

Indeed, in China the formula ‘cultural and creative industries’ has gained currency, linking creativity to heritage, artistic and customary cultures almost by definition. By contrast, there is a strand of work in Western creative-economy scholarship – including our own – that foregrounds innovation, dynamism, and entrepreneurial populism as the driving force. This means that there may be tensions between different definitions of creativity and culture (not just differences in what sectors are counted) across different jurisdictions. Such differences will produce widely varying and incommensurate datasets.

These issues of industry and sectoral definition, and also of the consistency of these classifications between local, state and national account, need to be carefully observed and reviewed. There are situations in which differing definitions may not pose a problem. If a Creative City Index is applied only to cities within a given jurisdiction (i.e. different cities within China), for instance, then Chinese definitions will apply consistently. However, if comparisons are sought across cities in different jurisdictions (say between China and Australia), then the results are only valid to the extent that the same definitional inputs are used.

A further point relates to the eighth dimension: openness, tolerance and diversity. These measures have been staples of creative city indexes since Richard Florida’s work over a decade ago. We include them here for the same reason: They provide a good measure of the tolerance to different lifestyles and beliefs, which is a good measure of a city’s willingness to tolerate and even embrace new ideas: it is a ‘proxy’ for openness to change and the readiness to value dynamism. But these issues can be difficult to measure because of sampling reluctance and reporting bias. The measures we have selected relate to levels of divorce (a proxy for women’s freedom), sexual openness (the ‘gay’ index) and religious tolerance. Sometimes official statistics on these matters can differ substantially from what is true, definitions can vary widely, and there may be official interference in or intolerance of data collection (to say nothing of the behaviours being measured). Hence these figures can be hard to track and verify accurately. These were the figures that we had the greatest trouble with in the pilot index (along with measures of local networks and interaction), and of which we are least confident. Nevertheless, they are included for good reason and we believe that the index’s quality is in significant part dependent upon quality inputs into this dimension.

We have also not sought to impress different weightings upon this index. Our initial supposition is that the ‘equal weightings’ model seems intuitively correct. Yet because some dimensions contained more variance than others, and also because the rankings were not always equal even in this small pilot sample, any change in the weightings will affect the overall outcome of the index and also the rankings.

Such issues will need to be taken up in future versions and applications of the CCI-CCI. We are confident that, with longitudinal comparisons as well as wider application to a greater number of cities, its accuracy as well as its utility will increase. It is like any system based on open connectivity; the more connections there are, the better it will be.