THE “CREATIVE CLASS” IN THE UK: AN INITIAL ANALYSISnaumannj/professional geography... · THE “CREATIVE CLASS” IN THE UK: AN INITIAL ANALYSIS by Nick Clifton CLIFTON, N. (2008):

THE “CREATIVE CLASS” IN THE UK: AN INITIAL ANALYSIS

© The author 2008Journal compilation © 2008 Swedish Society for Anthropology and Geography 63

THE “CREATIVE CLASS” IN THE UK:AN INITIAL ANALYSIS

byNick Clifton

CLIFTON, N. (2008): ‘The “creative class” in the UK: an initialanalysis’, Geogr. Ann. B 90 (1): 63–82.

ABSTRACT. Richard Florida argues that regional economic out-comes are tied to the underlying conditions that facilitate creativ-ity and diversity. Thus the Creative Class thesis suggests that theability to attract creativity and to be open to diverse groups of peo-ple of different ethnic, racial and lifestyle groups provides distinctadvantages to regions in generating innovations, growing and at-tracting high-technology industries, and spurring economicgrowth. In this paper we investigate the extent to which theremight be similar processes concerning the relationship betweencreativity, human capital, and high-technology industries at workin the UK as in North America. The approach taken is broadlysympathetic to the Creative Class thesis; critical perspectives andreservations from the literature are introduced as appropriate re-search is focused around the three principal research questions:Where is the creative class located in the UK? What is the impactof quality of place upon this dispersion? What is the connectionbetween the location of the creative class and inequalities in tech-nical and economic outcomes within the UK? To this end, the cre-ative class and its subgroups are defined and identified. We thenconstruct quality of place indicators relating to tolerance, diver-sity, creativity and cultural opportunity. To these are added meas-ures of public provision and social cohesion. Data are analysed bymeans of correlations and regression. In general we find that, al-though the distribution of the creative class is uneven and com-plex, our results are consistent with the findings of the NorthAmerican research with the notable exception of technology-based employment growth. Finally, priorities for further researchare discussed. The need to further investigate causality, variationswithin the creative class itself, and the potential role of qualitativedata in this are highlighted, as is the potential fate of “non-crea-tive” workers and places.

Key words: Creative Class, spatial analysis, economic dynamism,implications – policy and further research

IntroductionMuch of the recent interest in the development ofcreativity has drawn upon Richard Florida’s(2002b) book The Rise of the Creative Class.Whereas in the Industrial Age classical and neo-classical economic theory told us that ‘people fol-lowed jobs’, in the modern knowledge economyFlorida describes how ‘jobs follow talented peo-ple’. That is, places that display “creative class”characteristics, meaning a high presence of profes-sionals, technologists and bohemians, performedbest economically in recent years.

In a knowledge-based economy, the ability to at-tract and retain highly skilled labour is thereforeperceived as crucial to the current and future pros-perity of regions as well as entire nations. For ex-ample, Florida (2000) has argued that in the know-ledge economy, regions develop advantage basedon their ability to quickly mobilize the best people,resources and capabilities required to turn innova-tions into new business ideas and commercial prod-ucts. In particular, the ability to attract creative peo-ple in arts and culture fields and to be open to di-verse groups of people of different ethnic, racialand lifestyle groups provides distinct advantages toregions in generating innovations, growing and at-tracting high-technology industries, and spurringeconomic growth (Gertler et al. 2002).

This research demonstrates that quality of placemust be understood in broader terms than we havetraditionally been accustomed to: while the attrac-tiveness and condition of the natural environmentand built form are certainly important, so too is thepresence of a rich cultural scene and a high concen-tration of people working in cultural occupations(most specifically the “bohemians”). According tothe results from Florida et al.’s research the under-lying hypothesis is that the presence and concentra-tion of bohemians in an area creates an environmentor milieu that attracts other types of talented or highhuman capital individuals. The presence of such hu-man capital in turn attracts and generates innovative,technology-based industries (Florida 2002b).

However, given the interest Florida’s writingshave received from academics, policy-makers andthe media alike, it is no surprise that they have beenthe object of a high degree of critical examination.This critique has centred most notably around theapparent fuzziness of some of the concepts, defini-tions and causal logic Florida employs, the seem-ingly convenient appeal of his ideas to the agendasof a multitude of urban actors and policy-makers,and conversely the minimal attention paid to diffi-cult issues such as the potential inequalities andnegative externalities implied by a creative class

NICK CLIFTON

© The author 2008Journal compilation © 2008 Swedish Society for Anthropology and Geography64

model of regional development. More fundamen-tally, a number of authors question the very conceptof a “new economy” that can deliver prosperity intandem with greater levels of self-determination toan ever-expanding body of knowledge workers.Brown and Lauder (2006), for example, envisage afuture scenario of diminishing returns to humancapital investments, with all the various discontentsthis entails. Moreover, we also need to be awarethat “Culture” is now positioned at the centre ofmany urban policies. It has become a delivery ve-hicle for all manner of outcomes including socialcohesion, sustainability, economic growth, civicpride, mental and physical well-being, social inclu-sion, and an ever-increasing array of other social,economic and environmental goals. This trend ishighlighted by a recent report by the Department ofCulture, Media and Sport which states: ‘Culturedrives regeneration in many ways, from inspiringlandmark buildings through to reviving the decay-ing centres of market towns to bringing a commu-nity together around an arts event’ (DCMS 2004, p.6). In this climate we need to guard against the Cre-ative Class and the Creative and Cultural Industriesbecoming the latest policy panacea.

This paper essentially seeks to apply Florida’smodels systematically in the UK context,1 usingcomprehensive datasets, for the first time (to thebest of the author’s knowledge). The approach tak-en is broadly sympathetic to the Creative Class the-sis, although we are of course mindful of the criticalperspectives and reservations noted above. Thestructure of the paper is therefore as follows: thetheoretical basis of the Creative Class approach isoutlined in more detail below, with significant res-ervations and criticisms from the literature intro-duced as appropriate. Data sources and methodol-ogies are then described in some detail, with theconsequent results presented. These are focusedaround the three principal research questions:Where is the creative class located in the UK? Whatis the impact of quality of place upon this disper-sion? What is the connection between the locationof the creative class and inequalities in technicaland economic outcomes within the UK? Finally,conclusions are drawn and implications for futureresearch considered in the light of these questions.

The creative class: review of theoretical perspectivesA distinct advantage of city-regions is their abilityto produce, attract and retain those workers who

play the lead role in knowledge-intensive produc-tion and innovation – who provide the ideas, know-how, creativity and imagination so crucial to eco-nomic success. The idea that growth-based devel-opment agendas can be actively pursed at the citylevel is however not a new one – see, for example,the “urban entrepreneurialism” documented byLeitner (1990). If we accept that the value creationin many sectors of the economy rests increasinglyon non-tangible assets, the locational constraints ofearlier eras – for example, the access to good nat-ural harbours or proximity to raw materials andcheap energy sources – no longer exert the samepull they once did. Instead, what Florida and his as-sociates assert matters most now are those at-tributes and characteristics of particular places thatmake them attractive to potentially mobile, muchsought-after talent. A key reason for believing thata significant shift has occurred taking us into aknowledge economy is that data suggest this to betrue. Thus the book value of intangible assets com-pared to raw materials has shifted from 20:80 in the1950s to 70:30 in the 1990s (Cooke and De Lau-rentis 2002). Consequently, the distribution of tal-ent, or human capital, is an important factor in eco-nomic geography, as talent is a key intermediatevariable in attracting high-technology industriesand generating higher regional incomes. Thismakes it an important research task to explore fac-tors that attract talent and its effects on high-tech-nology industry and regional incomes (Florida2002c).

The replacement of raw materials or natural har-bours with human capital and creativity as the cru-cial wellspring of economic growth means that inorder to be successful in the emerging creative ageof the knowledge economy, regions must develop,attract and retain talented and creative people whogenerate innovations, develop technology-inten-sive industries and power economic growth. Suchtalented people are not spread equally across na-tions or places, but tend to concentrate within par-ticular city-regions. According to Florida, the mostsuccessful city-regions are the ones that have a so-cial environment which is open to creativity and di-versity of all sorts. The ability to attract creativepeople in arts and culture fields and to be open todiverse groups of people of different ethnic, racialand lifestyle groups provides distinct advantages toregions in generating innovations, growing and at-tracting high-technology industries, and spurringeconomic growth.

As Wojan et al. (2007) describe, theorizing on



how local environments influence economic out-comes has a long and rich history, the two dominantviews of which may be traced back to Marshall(1920) – agglomerations, industry/firm-focused –and Jacobs (1961, 1969) focusing on variety andpeople. Traditional theories of economic growthand development tended to emphasize the role ofnatural resources and physical assets. Such theorieswere used to inform strategies typically based onvarious incentives to try to alter the location deci-sion of firms. In recent years, several more relatedtheories have emerged. The first, associated withthe work of Porter (2000) and others, emphasizesthe role of clusters of related and supporting indus-tries. According to this work, clusters operate asgeographically concentrated collections of interre-lated firms in which local sophisticated and de-manding customers and strong competition withother firms in the same industry drive the innova-tion process. A second view associated with Lucas(1988) and Glaeser (1998) focuses on the role ofhuman capital – that is, primarily highly educatedpeople. It argues that places with higher levels ofhuman capital are more innovative and grow morerapidly and robustly over time. A third approach,associated with Richard Florida, emphasizes therole of creative capital, arguing that certain under-lying conditions of places, such as their ability toattract creative people and be open to diversity, in-form innovation and growth.2 Further independentresearch by Robert Cushing (2001) provides sup-port for the creative capital view. Peck (2005) re-marks upon how the Creative Class thesis taps intomany of the same ‘cultural circuits of capitalism’(Thrift 2001) as much of the immediately preced-ing work on the knowledge (or new) economy. Itshould also be noted that there is work precedingthe first accounts of the creative class which makesexplicit reference to quality of place and locationalchoice factors, including Wong (2001) in the UK,which in turn may be linked back to Hall et al.(1987).

Such constellations of talents and creative peo-ple are – as already mentioned – most commonlyfound in large city-regions where the diversity ofurbanization economies is most abundant. This, to-gether with other factors such as labour marketscharacterized by high demand for qualified person-nel, cultural diversity and tolerance, low entry bar-riers and high levels of urban service, largely de-termine the economic geography of talent and ofcreativity, both of which display concentration tolarge cities. According to Cooke and De Laurentis

(2002) cities on average are twice as advantaged bytheir knowledge intensity over towns and rural ar-eas compared to their already existing advantagesfrom agglomeration economies. Thus if a cityscores 50 per cent above the mean in GDP per cap-ita it is likely to score 100 per cent above it in termsof its knowledge-based industry. Thus there is morechance of knowledge economy employment in thecity than in the country, a major contributory factorin the renewed migration of young people from ru-ral to urban areas in many European countries,making the knowledge economy unevenly distri-buted and knowledge poverty a new kind of loca-tional disadvantage (Cooke and De Laurentis2002).

Thus, it is not enough to attract firms: the “right”people also need to be attracted. Richard Floridacalls for complementing policies for attractingfirms with policies for attracting people, whichmeans addressing issues of “people’s climate” aswell as of business climate (Florida 2002c). Indeed,the former is seen as basic to the latter, in that thepresence of human capital and talent is essential forattracting and developing high-tech industries andconsequently for the economic growth of cities.This suggests that the attention of politicians andplanners should be directed towards people, notcompanies; that is, away from business attraction totalent attraction and quality of place (Florida2002c). Critics of this viewpoint have howeverpointed to what they consider the “easy-sell” of thismessage to urban policy-makers; Peck (2005, p.752) in particular has dissected in some detail whathe regards as the eminently ‘deliverable supply-side policy prescriptions’ that flow implicitly fromacceptance of the Creative Class thesis, and whichthus (either by design or otherwise3) find a readymarket among these newly “legitimized” urban ac-tors.

The knowledge-based economy means then thatthe ability to attract and retain highly skilled labouris seen as crucial in terms of both the current and fu-ture prosperity of city-regions as well as entire na-tions. Recent research on this question indicatesthat talent is attracted to and retained by cities, butnot just any cities. In their analysis of Americanmetropolitan areas, Richard Florida and GaryGates have shed new light on those characteristicsof urban regions that seem to be most important inthis process (Florida and Gates 2001; Florida2002a, 2002b, 2002c). The central finding of thiswork is that the social character of city-regions hasa very large influence over their economic success

NICK CLIFTON


and competitiveness. In particular, Florida and hiscolleagues have found that those places which offera high quality of life and best accommodate diver-sity enjoy the greatest success in talent attraction/retention and in the growth of their technology-in-tensive economic activities.

Diversity as a key aspect of successful placesconcerns entry barriers facing newcomers: citieswith great diversity are understood as places wherepeople from different backgrounds can easily fit in(Florida 2002c). There are several quantitative in-dicators used by Florida and colleagues to capturethis. One of the most influential variables wasfound to be a city’s “gay index”, measuring theprevalence of gay males in the local population.This index has been shown to imply a high degreeof openness to newcomers of diverse backgroundswith respect to nationality, race, ethnicity and sex-ual orientation, reflecting an environment that isopen to diversity, high in urban oriented amenities,and characterized by low entry barriers (Florida2002c). Another associated indicator of diversity isthe “melting-pot index”, reflecting the proportionof a city-region’s population that is foreign-born;this is the indicator we employ in the present study.

The findings from the Canadian study by Ger-tler et al. (2002) strongly indicate that the relation-ships first captured for US city-regions in the workof Florida and colleagues are also evident in Ca-nadian city-regions. If anything, the relationshipsin Canada are stronger than those found in theUnited States. In particular, the findings showedthat a vibrant local creative class and an opennessto diversity attracted knowledge workers in On-tario and Canada more generally. It was also foundthat, in general, Ontario city-regions (of which To-ronto is the largest) have a solid foundation inthese areas to compete against US city-regions. Insummary, there appears to be a strong set of link-ages between creativity, diversity, talent and tech-nology-intensive activity that are driving theeconomies of Ontario’s – and Canada’s – city-re-gions (Gertler et al. 2002).

Creativity, class and the economic mainstreamThe creative class moniker can be linked more gen-erally to a reassessment of the idea that “bourgeois”values of business, profit and so on (inherentlylinked to a wider conservative value system) are bydefinition mutually exclusive to “bohemian” valuesof creativity, the embracing of new ideas and valu-ing of diversity (in its various forms). Brooks

(2000) actually merged the two words themselvesin describing the emergence of the bobos – a newgroup of people in which bourgeois and bohemianvalues are blended into the creative, unconvention-al but also entrepreneurial (Brooks’ depiction isthough an inherently less sympathetic one thanFlorida’s description of his creative class). Thisidea has some intrinsic appeal when the nature of“cool” jobs in technology, new media and so on isconsidered: as Heath and Potter (2006, p. 206) putit: ‘What people yearn for these days is no longeran old-fashioned “status” job, like being a Doctor.The “cool job” has become the holy grail of themodern economy.’ Indeed, with play-zones andchill-out areas having become the stuff of hi-techstart-up cliché the question remains as to whethersuch idiosyncrasies are actually intrinsic to thiskind of creativity or merely ostentations, to bedropped hastily in an economic downturn such asthe fallout from the dot.com shakeout. A full an-swer to this question is however beyond the scopeof this paper.

The use of the word class comes of course witha certain baggage in that it implies some kind ofself-identity and consistent value system within asocio-political hierarchy. Whether the creativeclass really is a class in some kind of Marxist senseis something of a moot point; the broad attitudesheld and approaches to life that Florida (and others)describe does suggest that at least the term is notwholly erroneous in this context, but this argumentis not an entirely convincing one. These traits in-clude personal attire and style, beliefs and values,attitudes to work – to old-style demarcations ofblue collar and white collar is added the “no-collar”workplace; viewed through the eyes of someoneused to the traditional demarcations of the work-place these are people ‘who seem to be alwaysworking, and yet never working when they are sup-posed to’ (Florida 2002b, p. 5).

The points that Markusen (2006) makes regard-ing the fuzziness of Florida’s definitions and caus-al logic are thus warranted in certain respects, al-beit that the comment ‘it is rather amusing to thinkof the vast bulk of artists as making common urbanor economic cause with bankers, real estate devel-opers’ (Markusen 2006, p. 1937) does seem ratherclose to something of a straw man. Artists (i.e. thebohemians) are typically treated as a distinctgroup even within Florida’s creative class. Indeedthe bohemians are essentially posited as a qualityof place factor in their own right; that is, one whichis attractive to the much wider creative class in



general, which does not necessarily imply a com-plete commonality of values and beliefs.4 Further-more, the ingenious study by Wojan et al. (2007)did find evidence for a positive “artistic milieu”effect upon regional growth outcomes; these au-thors attempt to avoid the potential bohemians-lead-to-growth, growth-leads-to-bohemians cir-cularity by controlling for both supply and de-mand-side factors for artists, and in turn using theresiduals from this model (i.e. unexplained varia-tions in the presence of bohemians) to predict eco-nomic growth. The positive result does not ofcourse necessarily mean that artists themselvesare the conscious hard-nosed economic drivers ofthe localities in which they find themselves con-centrated. Either way, a conceptual route in tothese issues of shared versus divergent preferenc-es is offered by an analysis of the knowledge baseswhich differentiate occupations within an aggre-gate creative class (Hansen et al. 2005). We are notable to emulate that here, but we can at least spec-ify separate models for the Creative Core, Crea-tive Professionals and combined Creative Class,and indeed omit the bohemians from the analysiswhen they are also involved as an independentquality of place variable.

However, even when the creative class is definedin its widest sense, this still implies that around 60per cent of the workforce is engaged in “non-crea-tive” activities. A number of authors make the pointthat the role of this non-creative class is neglected.Peck (2005) goes so far as to suggest that Floridaadvocates a form of what he (Peck) terms ‘creativetrickle-down’; that is, there is an acknowledgementof the potential inequalities and negative external-ities associated with high-growth locations, butvery little in the way of concrete policies for howthese might be addressed, and generally nothingwithin the remit of central government. There hasbeen some attempt to deal with these concerns out-side the US context (Cannon et al. (2003) in theUK; Bradford (2004) in Canada) within some ofthe creative-cities literature, but typically theseamount to little more than somewhat ad hoc socialobjectives bolted on to an underlying CreativeClass orthodoxy. Whatever the subtleties of the de-bate however, the fundamental point remains that adichotomous split between the bohemian and thebourgeois, or the business like and the creative, isno longer adequate in describing how a significantgroup of people live and work, and that this creativeclass is now very much part of the economic main-stream.

Research questionsAlthough North America and Europe share manycommon values and institutions, there are aspectsof their respective societal development that showstrong divergence; for example, with regard to po-litical priorities, functioning of labour markets,economic growth processes and social outcomes(see e.g. Hall and Soskice 2001). Given that virtu-ally all of the systematic research of the phenome-non has taken place in North America, the CreativeClass thesis warrants investigation in a Europeancontext.5 Therefore, the research reported in thispaper represents an analysis of quality of place andthe dispersion of the Creative Class in the UK,building upon the work described above that hasbeen undertaken in North American cities in orderto understand whether similar processes concern-ing the relationship between creativity, human cap-ital and high-technology industries are at work inEurope as clamed within North America. The ques-tions we seek to address are then essentially:

1 Where is the Creative Class located in the UK?2 What is the impact of quality of place upon this

dispersion? Does the Creative Class thesis ap-pear valid within the UK context?

3 What is the connection between the location ofthe Creative Class and inequalities in technicaland economic outcomes within the UK?

Our research is taking place within the context of awider project entitled ‘Technology, talent and tol-erance in European cities: a comparative analysis’,involving matched datasets and research partners inDenmark, Sweden, Norway, Finland, the Nether-lands and Germany.6 To get at these issues, the roleof human capital, creative capital and diversity intechnology-based economic development in theUK is investigated. The research uses the twomeasures developed by the North American stud-ies: the Bohemian index to reflect creative capital,and the Diversity (mosaic) index to reflect open-ness and diversity. This suggests that there will bea relationship between openness to creativity anddiversity and the ability to support high-tech indus-tries and economic development based on talentedworkers. New indices are developed in order tograsp the fundamental differences in certain as-pects of life between North American and Europe-an societies (see the following section). As noted,this kind of analysis has not yet been performed forEuropean cities, and has the potential to shed im-

NICK CLIFTON


portant new light on the role of quality of place inshaping the competitiveness of city-regions in theUK.

Data and methodologyThe data for the quantitative statistical analyses arederived primarily from the 2001 Census of Popu-lation, supplemented by the Labour Force Surveyand the Annual Business Inquiry. Various indicesare constructed from these data as described below,and explored through means of plots, regressionsand correlations. In addition, this is being supple-mented by the collection of qualitative empiricalmaterial in the form of interviews with key actors,as well as with theoretical representative groups of“talents” employed in urban governance and plan-ning in the case study cities, high-technology in-dustry and service, and higher education and re-search institutions. These interviews are beingcomplemented with additional interviews with rep-resentatives from creative/artistic occupations aswell as from different ethnic groups. The qualita-tive analyses are aimed at obtaining a subjectiveevaluation and assessment of the relative impor-tance of the various indices used in the quantitativeanalyses in order to obtain a more comprehensivepicture of which preferences talented people actu-ally have and why they behave the way they do.These interviews are ongoing at the time of writing;thus most of the material presented here derivesfrom the quantitative analyses.

Key variables for quantitative analysesThe key variables for quantitative analyses are theBohemian index, the Talent index, the Diversity in-dex, and the Tech-Pole index. These mirror varia-bles were employed in previous research by Florida(2002a, 2002b, 2002c) and Gertler et al. (2002) onthe geography of talent and the rise of the creativeclass. In addition, indicators for cultural and recre-ational amenities, which were also used in Flori-da’s studies, will be considered. A pair of new in-dicators, developed to reflect characteristics of Eu-ropean cities and their national political econo-mies, is also introduced. These two new indicatorsare social cohesion, and a public provision indexmeasuring the supply of public sector goods suchas education, healthcare, social security and so on.

In general, the variables used were designed tojointly maximize consistency between the differentEuropean countries involved in the wider project,

and between Europe and the USA and Canada. Var-iation in availability across the partner nations hasinevitably imposed constraints upon the data usedby individual partners, and in some cases involvedcompromises around lowest common denominatorlevels of detail and time frames.

Creative class: Clearly of central importance is theability to actually quantify the size of the creativeclass present in any given location. These are es-sentially people who as a key constituent of theirwork are involved in the creation of new know-ledge, or use of existing knowledge in new ways,combinations and so on. In the absence of a primarydataset relating to the actual engagement in suchactivities, this is proxied by the use of occupationalcategories. We subdivide the Creative Class into theCreative Core (scientists and engineers, architectsand designers, academics and teaching profession-als),7 and the Creative Professionals (associatedprofessional and technical occupations of the Cre-ative Core, managers, financial and legal profes-sionals). Data for this index (and the others usingoccupational data) were derived from the 2001Census of Population, the only source of sufficient-ly similar size to allow four digit Standard Occu-pational Classification 2000 (SOC2000) break-downs at the levels of regional disaggregation re-quired.

Bohemian index: The Bohemian index is definedusing employment in artistic and creative occupa-tions. It is a locational quotient that compares theregion’s share of the nation’s bohemians to the re-gion’s share of the nation’s population (local prom-inence of employment in artistic/creative occupa-tions compared to national prominence of employ-ment in the same occupations). Data for the UK aredelineated using the SOC2000 system; within thisit is convenient to use sub-major group 34 (Culture,Media and Sports) to define the bohemian occupa-tions.8

Cultural Opportunity index: The opportunity to en-joy cultural and recreational activities can be cal-culated more directly than other indicators on qual-ity of place. According to Florida, such opportuni-ties play an important role in cities’ ability to attractthe creative class. The proportion of employees inthe cultural and recreational industries within anarea is used as an indicator of this cultural oppor-tunity. This involves using a number of three digitStandard Industry Classification (SIC) groups from



Annual Business Inquiry to account for employ-ment within restaurants and bars, libraries, muse-ums and other cultural and entertainment activities.

Talent index: Talent is defined as the proportion ofthe population over 18 years of age with a Bach-elor’s degree or higher (local proportion comparedto national). These data were obtained from the2001 Census of Population.

Diversity (Openness) index: The Diversity index isthe counterpart of Florida’s melting pot index. It iscalculated as the proportion of the total populationthat is foreign-born. A second index, which meas-ured the proportion of total population that is for-eign-born of non-Western origin, was also con-structed; the logic behind this was that a more vis-ible manifestation of diversity may be a more validtool, or at least provide greater insight into the con-cept of tolerance (see later discussion of this issue).These data were provided by the 2001 Census ofPopulation.

Tech-Pole index: Following the method of theMilken Institute (e.g. DeVol et al. 2007), this indexshows local employment in technology-intensivemanufacturing and service sectors (specializationand size). The index compares a region’s share ofnational employment in high-technology indus-tries to the region’s overall share of national em-ployment; this is then adjusted for city size by mul-tiplying by a region’s overall share of national high-technology employment. Therefore, it reflects boththe region’s degree of specialization in technology-intensive activity as well as its sheer scale of em-ployment in these sectors. The index includes tech-nology-intensive sectors in both manufacturingand services. Data for constructing the index wereextracted from the Annual Business Inquiry (ABI)employee analysis for 2002.

Social Cohesion index: Florida himself has ex-pressed concern regarding the polarization in termsof prosperity that high levels of creative class em-ployment has been associated with in the USA, tothe point where this is becoming a negative qualityof place factor for these creative workers them-selves. As such it was felt that some attempt to fac-tor this into the research design should be made;that is, not cohesion in the sense of a social capitalindicator (which may indeed be incompatible withthe concept of openness), but rather one based onperceived inequality. Ideally this should involve

some kind of gini coefficient with respect to in-come distribution, but these kinds of data are notavailable at the spatial level we are using (this isalso true for the other European project partners).As such, the apparent level of exclusion from main-stream economic activity was used, as representedby the ILO definition of unemployment.9 Using alabour market variable as a proxy for social cohe-sion introduces a new net of complicating factors –as with other issues relating to the operationaliza-tion of Creative Class concepts, these are revisitedin later sections. The data were taken from the La-bour Force Survey with respect to the twelve-month period March 2001 to February 2002.

Public Provision index: Having acknowledged thepotential implications for social cohesion impliedwithin the Creative Class model, there is a need tofurther take account of the European context; giventhe high levels of private employment-based provi-sion in the USA, particularly for healthcare, to whatextent might high-quality social welfare provisionbe a quality of place factor for the European crea-tive class? In this index, levels of employment ineducation and healthcare (SIC2003, two-digitcodes 80 and 85) for a given locality are expressedas a proportion of the resident population, usingdata from the Annual Business Inquiry.

Levels of geography: As described above, data werecollected such that the triple goals of country-spe-cific, inter-country and Europe–North Americacomparisons could best be achieved. This was alsotrue with respect to the levels of geography (i.e. spa-tial units employed in the analysis). With the partnercountries accounting for large variations in size,governance structure, patterns of population disper-sion and so on it was impractical to impose a singlestandard definition; in practice such definitions ac-tually have different meanings dependent upon thecontext in which they are applied; i.e. the same def-inition will actually define a spatial unit that has adifferent meaning from one country to another.

As we were seeking definitions that encapsulat-ed something approaching functional labour mar-kets (analogous at least in part to the municipalcity-regions employed in the North Americanresearch10) it was decided that the most meaningfulfunctional unit in each national context would beused, subject to this also being a level at which thenecessary statistical data were available from therelevant national agencies. As Parr (2005) notes,standard UK administrative geographies do not

NICK CLIFTON


typically relate in a systematic way to any theoret-ical construct of the city-region; for the UK thismeaningful unit was primarily the NUTS3 defini-tion (105 spatial units in England and Wales). Dueto the sometimes complex and non-hierarchical na-ture of these standard geographies, this is supple-mented by analysis using the Unitary/County Au-thority level (171 units).

A significant difference between our geographyand that of the research undertaken in the USA andCanada is that it is totally inclusive (i.e. no locali-ties are excluded), whereas the North Americananalysis has tended to focus exclusively on largeurban centres. This was essentially done for rea-sons of consistency; the European countries aregenerally more densely populated than NorthAmerica, the demarcation of city-regions less ob-vious and so partial selection is potentially more ar-bitrary. Once information has been included in adatabase it can always be manipulated further, re-scaled or partially excluded, while the reverse ismuch more problematic.

Results and analysisMapping the creative class in England and WalesAs shown in Table 1, the creative class in Englandand Wales accounts for some 37.3 per cent of theworkforce, substantially greater than the ‘morethan 30%’ figure that Florida (2002b) himselfquotes with regard to the USA. Problems in obtain-ing consistent occupational time series data meanthat it is difficult to draw many conclusions with re-gard to how the size of the creative class may bechanging over time. However, if the major groupProfessional Occupations is taken as a proxy for theCreative Core, then an increase is observed frombelow 9 per cent of the workforce in the 1991 Cen-sus to over 11 per cent in 2001, suggesting signif-icant growth in these occupations.

The total figure for the Creative Class is split be-tween the Creative Core (9.7% of the workforce),the Creative Professionals (25.5%) and the Bohe-mians (2.1%). As Table 1 shows there is consider-

able variation around the England and Wales per-centages. It is worth noting here that the twoNUTS3 areas of Inner London West and Inner Lon-don East between them account for the highest per-centages across all four creative class categoriesshown. Other than Inner London West (64.9%),only Inner London East (at just under 52%) pos-sesses a labour force of which the Creative Classcomprises more than half. For the bohemians aneven greater concentration is observed; after thetwo Inner London areas (both above 7%) the high-est percentage is found in Brighton and Hove at 4.4per cent of the labour force.

At the opposite end of the spectrum, the same lo-calities also tend to crop up across the board, albeitwith some variation; Stoke on Trent possesses thelowest proportion of total Creative Class and alsoCreative Core, Gwent Valleys the lowest share ofbohemian occupations, while the lowest share ofCreative Professionals is found in Hull. Despitethis apparent variation, the same localities are typ-ically found within a few places of each other atboth the top and bottom of the rankings. This ef-fectively demonstrates that although there are var-iations in the overall make-up of the Creative Classin any given place, there is little correlation be-tween this variation and the actual size of the Cre-ative Class in that place.

Table 2 shows the top fifteen and bottom fifteenNUTS3 areas in England and Wales ranked interms of their population size. The actual popula-tions vary from just under 1.8m down to around69 000, with a median size of just over 360 000. Thecorresponding ranks in terms of Creative Class Lo-cation Quotient (LQ) are also shown. The LQ itselfis a measure of spatial concentration, expressed asa proportion such that the average for England andWales is 1.

These data highlight the sometimes arbitrary na-ture of NUTS3 geography in the UK, with counties,towns, cities and urban metropolitan areas all beingrepresented. Many of the smaller areas are partic-ular accidents of geography such as islands and rel-atively isolated smaller towns and cities.

Table 1. Creative Class as a share of the labour force (%).

England and Wales Highest locality Lowest locality

Creative Core 9.7 13.7 5.0Creative Professionals 25.5 44.3 17.7Bohemians 2.1 8.8 0.8Creative Class in total 37.3 64.9 24.1

Source: Census of Population (2001).



One thing which Table 2 does highlight are thedifferences that exist within London, particularlythe relatively low ranking in terms of CreativeClass LQs of the more peripheral areas in contrastto the central areas. Although London looms largeover the UK picture, there are other regional cen-tres of high creativity (and high concentrations ofbohemians). The situation tends to differ in thesmaller countries, where the capital may be theonly realistic locational choice for many special-ized professional workers. Moreover, the status ofLondon as a genuine global city is significant, as itis therefore competing for creativity (to use thefashionable rhetoric) on the world stage, with thedifferent tensions that this brings. One of these maybe a latent conflict between national capital versusinternational roles, perhaps generating competingpressures or funding priorities; for example, re-sources which are ostensibly earmarked for localregeneration might be spent with one eye on the po-tential national or international payoffs.

As the UK’s “second city”, Birmingham is an in-teresting case; with statistics relating to the centralarea (population of about 1m) it ranks proportion-ally much lower for the Creative Class (54th) thanit does in terms of population (13th); variationsaround this urban core are however apparent in Fig.1, discussed below. Despite these discrepanciesand data issues, there is some evidence of an asso-ciation between size of agglomeration and creativeclass concentration, a rank correlation between thetwo producing a coefficient of 0.41. In order to ex-plore this agglomeration effect further, the associ-ation between population density and CreativeClass concentration was investigated; somewhatsurprisingly, virtually no link was found, with acorrelation coefficient of only 0.1.11 Clearly then itis not just size or density of population, the higherlevels of which are typically found in metropolitanareas and urban centres, that is associated with thelocation of the creative class. This could reflect dif-ferent preferences within the agglomerated crea-

Table 2. Creative Class location and population rank.

Population size Creative class LQrank Locality (NUTS3) rank

1 London (inner, west) 12 London (outer, west and N west) 43 London (outer, east and N east) 294 Kent 365 Greater Manchester (south) 356 Essex 257 Hampshire 178 Greater Manchester (north) 759 London (outer, south) 6

10 Lancashire 5011 Surrey 312 Hertfordshire 913 Birmingham 5414 London (inner, east) 215 Calder, Kirlees and Wakefield 66

91 York 3292 Swindon 4793 Herefordshire 5194 Southend 2695 Telford and Wrekin 7396 Peterborough 5597 Thurrock 9798 Blackpool 8699 Blackburn and Darwen 87

100 Isle of Wight 65101 Torbay 68102 Powys 80103 Gwynedd 70104 Darlington 56105 Anglesey 76

Source: Census of Population (2001).

NICK CLIFTON


tive class, as suggested by some of the literature re-viewed earlier. It also suggests that different spatiallevels of analysis might be appropriate – in order,for example, to zoom in on micro-level neighbour-hood effects not currently captured in the data.More generally, these issues highlight the impor-tance of investigating additional quality of placefactors (quantitative and if possible qualitative) inseeking to explain the distribution of the creativeclass; see the section on the creative class and qual-ity of place below.

Table 3 provides a slightly different view of Cre-ative Class distribution, in this case showing the tenhighest and lowest unitary/county authorities,ranked with respect to their Creative Core LQs.There are in total 171 of these localities in Englandand Wales, and this level of geography allows a lit-tle more detail to become apparent than is the casewith NUTS3. Further to this end, the thirty-threeunitary authorities that comprise London have beencombined into the single standard NUTS1 UK re-gion; this provides a view of how London as awhole is positioned, but equally important is thefact that eighteen of the top twenty Creative CoreUAs in England and Wales are located in the capi-tal,12 and so collapsing these into a single figure al-lows detail elsewhere to emerge.

As might be expected (see e.g. Hall et al. 1987),localities in the west-of-London M4 corridor area(Wokingham, Reading, Oxfordshire, Windsor andMaidenhead) feature heavily in the top ten CreativeCore LQs. In addition to London, ranked at numberseven Cambridgeshire completes the third facet ofthe “Golden Triangle” of the UK’s knowledgeeconomy. What is perhaps more interesting is thatin addition to those areas which might be expectedto feature, a number of less obvious regional cen-tres of creativity emerge – Cardiff in the west, Man-

chester in the northwest (Trafford lying just to thewest of the city centre with Manchester itselfranked only four places below at 14), and Newcas-tle in the northeast. Finally, our rankings confirmthe perception of Brighton and Hove as a creativecentre with its unique bohemian image and relativeproximity to London.13

Turning attention to the bottom ten UAs, anumber of these are places suffering the protractedafter-effects of the loss of heavy industry, either asdistinct localities (Blaenau Gwent, Stoke on Trent,Barnsley) or the deindustrialized areas of large cit-ies, for example, Tameside (Manchester), Knows-ley (Liverpool) and Sandwell (Birmingham). In ad-dition, a couple of places associated with old-styleseaside holiday resorts also make an appearance(northeast Lincolnshire, Blackpool). Although lo-cations in the left-hand column are generally asso-ciated with higher levels of growth and prosperity,the patterns revealed here do not necessarily implyan economic consequence; for example, Blackpoolalthough the ninth least Creative (Core) locality inEngland and Wales has had some success in recentyears, reinventing itself from a traditional “bucketand spade” destination into one that seeks to attracta younger party-oriented clientele.

Figure 1 shows how the Creative Core are dis-tributed within England and Wales, with Londonexcluded from this representation for the reasonsoutlined above. A number of locations are high-lighted on the map; this is not necessarily a com-prehensive listing of the highest or lowest rankedplaces, rather they are intended to serve as illustra-tive examples. The concentration Creative Core inthe southeast of England generally and the M4 cor-ridor area in particular is apparent; within this areathe Unitary Authority of Reading is highlighted,immediately to the southeast is Wokingham, and

Table 3. Creative Core location by unitary authority/county.

Top 10 localities (LQ) Bottom 10 localities (LQ)

1. Wokingham 1.46 1. Barnsley 0.632. Reading 1.42 2. Tameside 0.623. Cardiff 1.39 3. NE Lincolnshire 0.614. Oxfordshire 1.34 4. Knowsley 0.605. London* 1.33 5. Kingston upon Hull 0.586. Newcastle 1.32 6. Sandwell 0.577. Cambridgeshire 1.31 7. Thurrock 0.568. Brighton and Hove 1.31 8. Blackpool 0.529. Windsor and Maidenhead 1.31 9. Blaenau Gwent 0.51

10. Trafford 1.27 10. Stoke on Trent 0.49

Source: Census of Population (2001).Note: * Combined NUTS1 region.



Windsor and Maidenhead lie to the east again (notshown).

The three cities of the East Midlands (Derby,Leicester and Nottingham) serve to illustrate theway in which urban centres do not necessarily pos-sess a consistent relationship with their surround-ing hinterland in terms of where the Creative Corecongregate – one of these cities (Leicester) has asignificantly higher concentration than the corre-sponding county area, while in the other two casesit is approximately equal. There is a wider point tonote on this in that as alluded to earlier, explana-tions of creative class distribution go beyond a sim-plistic urban vs. rural issue, although this may playsome part in the process. It is also worth acknowl-edging that such idiosyncrasies do tend to get lostin a NUTS3 level or user-constructed city-region-type analysis.

This type of pattern is further revealed when themajor cities are examined, in that localities whichmake up the wider city area are often very differentwith respect to levels of the Creative Core, despitequite small distances being involved. Examples of

this are apparent in Birmingham, with the district ofSandwell immediately to the west placed fifth low-est (see Table 3) while directly to the east lies Soli-hull (twenty-second from the 138 Unitary Authori-ties and Counties outside London). A similar situa-tion is observed with respect to Manchester: Tame-side to the east, Trafford to the (south) west, placedin the bottom ten and top ten respectively of the Cre-ative Core ranking. The northeast of England pro-vides another example of contrasting creative classpatterns existing in close proximity; this is howeverdifferent in one key aspect in that although adjacentto one another, Newcastle and Sunderland are twodistinct cities with their own identities, not to men-tion rivalries. Newcastle’s much higher levels ofCreative Core may in part be assigned to its role asa regional capital and administrative centre, but inrecent years it has seen extensive cultural develop-ment and has acquired something of a boho imagethat is more complex to explain.

Focusing on those areas of England and Walesthat are less well placed in terms of creativity, asdiscussed above a number of larger city areas fall

Fig. 1. Location of the CreativeCore in England and Wales (exceptLondon).

NICK CLIFTON


into this category, along with distinct ‘post-indus-trial’ regions such as the South Wales valleys, andthe former coalfield areas in the north of England.It is notable that the former exist in relative prox-imity to Cardiff, one of the highly placed cities forCreative Core in England and Wales, and it is sig-nificant that of these Blaenau Gwent is one of thetwo valley areas that does not share a border witheither Cardiff or Newport to the east (the other be-ing Merthyr Tydfil).

The creative class and quality of placeAs outlined earlier, Florida suggests that the crea-tive class is highly mobile, with strong preferencesfor certain aspects of quality of place. In conjunc-tion with Canadian researchers (see e.g. Gertler etal. 2002) he shows that in North America citieswith high levels of creative class tend to be open,tolerant and diverse places, with high levels of rec-reational and cultural opportunity. According toFlorida (2002b) an open environment is one inwhich people are accepted and allowed in, on thebasis of their skills rather than their similarity to theexisting gatekeepers. As such, another indicator oftolerance we can employ is the proportion of bohe-mians in any given location. This group oftenstands out compared to mainstream culture, life-style and values, and thus in order to thrive requireshigh levels of tolerance (Florida 2002b, pp. 260–261). The idea therefore is that bohemians seek theopportunity to experience a diversity of impres-sions, and are thereby themselves indicators of atolerant and open environment (it may be recalledfrom previous sections that this view has been cri-tiqued as revealing a somewhat simplistic under-standing of the locational choices of arts- and cul-ture-based employment; e.g. Markusen 2006; Wo-jan et al. 2007). Finally, to these indicators havebeen added those of Public Provision and Unem-ployment (Social Cohesion) in order to reflect theEuropean context of our research.

We do not suggest that all aspects of a conceptas nebulous as quality of place can be perfectly cap-tured by these relatively simplistic statistics; it ispossible though to construct some general indica-tors, underpinned by Creative Class theory – in thissection we analyse the association between theseindicators and the location of the creative class inEngland and Wales. An analysis is conducted of thebivariate correlations between the individual qual-ity of place indicators and the location of the crea-tive class; we then combine these indicators into asingle multiple regression model, which allows es-timation of the overall explanatory value of thesevariables on the distribution of the creative class.

Table 4 shows summary values of the quality ofplace indicators for England and Wales. The distri-bution of the bohemians has been discussed aboveand so will not be commented upon here. Levels ofdiversity, defined as the percentage of residents thatare foreign-born, shows quite considerable varia-tion; the three highest values are accounted for byLondon NUTS3 areas (of which there are five) withInner London East being the highest. The highestplaced non-London locality is Leicester, a city witha long tradition of immigration from the Indiansubcontinent. This result does raise concerns overthis particular indicator, and these are subsequentlydiscussed. Conversely, the Gwent Valleys is theleast diverse locality in England and Wales by thismeasure, closely followed by the other Valleys area(central).

Cultural opportunity also sees wide variations,however, Inner London west is a massive outlier,unsurprisingly, given the concentration of high pro-file museums and galleries therein. Perhaps moresurprising is that the next highest placed locality isBlackpool at 4.1 per cent; again this highlights howcertain quality of place indicators can be influencedby underlying factors which are not necessarilyconsistent – the Blackpool figure is almost certain-ly largely derived from a high concentration of barsand amusements rather than the “high” culture

Table 4. Quality of place indicators: overview (%).

England and Wales Highest locality Lowest locality

Openness (diversity) 8.9 36.2 1.6Bohemians 2.1 8.8 0.8Cultural opportunity 2.9 15.2 1.3Public provision 9.0 17.6 5.4Unemployment 5.1 10.5 2.1

Source: Census of Population (2001); Annual Business Inquiry, Employee Analysis (2002).



found in central London. Once again, the lowestvalue of the indicator (1.3%) is accounted for by theGwent Valleys region.

Finally, with regard to the Public Provision in-dex (PPI) it is quite hard to discern any patternwithin the results, with two of the top three beingcities of the East Midlands (Nottingham andLeicester respectively), separated only by InnerLondon west. Again, different forces are likely tobe at work here.

As shown in Table 5, the indicators for both di-versity and the bohemians are positively correlatedwith the localization of the Creative Class and bothof its subgroups. This means that the creative classin England and Wales tends to live in places thatalso have high levels of bohemians and diversity.Both relationships are quite strong, particularly sobetween the creative class and the location of thebohemians. The Openness index is a fairly simplis-tic measure of tolerance and as such might not besufficient. A more focused measure on the effect of

integration, or for highly educated foreign-bornworkers might, capture the openness of a commu-nity better. These are however restricted by dataproblems, an issue that forthcoming qualitativework will seek to address. However, from the abovewe can tentatively conclude that the Creative Classand tolerance (measured as diversity, and the pres-ence of bohemians) do correlate in the same waythat Florida and his associates found in the NorthAmerican analysis. Figures 2 and 3 show graphicalrepresentations of the relationship between theCreative Class and respectively the bohemians andopenness. Figure 2 demonstrates a strong and pos-itive relationship as predicted, with the fourNUTS3 areas of London, along with Brighton andHove as significant outliers (i.e. they have higherconcentrations of bohemians for their observedlevels of Creative Core employment than is typi-cal). This makes intrinsic sense given the locationsin question, but without additional data we can onlyspeculate on the relative importance of supply ver-

Table 5. Quality of place – bivariate correlations.

Correlation with Correlation with Correlation withCreative Class Creative Core Creative Professionals

Openness (diversity) 0.52** 0.44** 0.50**Bohemians 0.72** 0.58** 0.71**Cultural opportunity 0.52** 0.29** 0.56**Public provision 0.02 0.21* –0.07Unemployment –0.31** –0.21* –0.33**

Source: Census of Population (2001), Annual Business Inquiry, Employee Analysis (2002).Notes:* Significant at the 95% level.** Significant at the 99% level.

Fig. 2. Creative Core and Bohemi-ans (Boho LQ).

NICK CLIFTON


sus demand factors within this pattern. Figure 3 il-lustrates the point made earlier that despite a sig-nificant trend line, localities can possess identicallevels of openness (as we measure it), but very dif-ferent levels of the Creative Core. Different localand regional trajectories (and thus potentially dif-ferent relationships) are likely to underpin this ob-servation.

With regard to cultural opportunity, a positiveand statistically significant relationship exists be-tween the presence of the creative class in a loca-tion, and that location’s cultural and recreationaloffering. This finding is again compatible withthose of the North American research. This rela-tionship is quite strong for the Creative Class as awhole and for the Creative Professionals, but some-what puzzlingly weaker for the Creative Core.

The concepts of public provision and social co-hesion are actually quite abstract, and are thereforesomewhat difficult to operationalize in practice.The relative provision of public goods and servicesin healthcare and education is measured here by

levels of employment in these sectors. Conversely,unemployment (i.e. exclusion from the labour mar-ket) is one of the main manifestations of social ex-clusion. For public provision, a significant and pos-itive (but weak) association is observed for the Cre-ative Core only. For unemployment, the relation-ship is consistently negative, although moderateand weak with respect to the Creative Core, mean-ing that municipalities with high levels of unem-ployment tend to have a low concentration of thecreative class.

Overall, it would appear that the creative class inEngland and Wales shows a similar pattern of dis-tribution with respect to quality of place, as is ob-served in North American cities. High concentra-tions are typically found in places which are toler-ant, diverse, bohemian, socially cohesive andwhich offer higher levels of cultural opportunity,but the overall pattern is quite complex.

In order to test a unified Creative Class model,the multiple regression method is used. The ad-vantage of using this method is that all indicators

Fig. 3. Creative Core and Open-ness (diversity).

Table 6. Results of the combined model: summary.

Total Creative Class Creative Core Creative Professionals

Openness (diversity) + n/s +Bohemians + + +Cultural opportunity n/s – +Public provision n/s + –Unemployment – – –

Adjusted R2 .664 .484 .698



are tested in one model, and therefore it is possibleto control for any multicollinearity between theindependent variables. A total of six models wereproduced; for each of the Creative Class and thetwo subgroups (the Creative Core and the CreativeProfessionals) two models were calculated foreach dependent variable, including either all for-eign-born citizens or non-Western foreign-borncitizens as the Openness variable. These resultsare summarized in Table 6: this reveals some var-iation across the regressions: the model for theCreative Core had the lowest explanatory power(.484), while that of the Creative Professionalshad the highest (.698). There was also some vari-ation across the quality of place variables: for ex-ample the Cultural Opportunity was negatively as-sociated with the Creative Core (the opposite di-rection predicted), and not significant within theTotal Creative Class model. The Public Provisionindex was also not significant for the Total model,but positively associated with the Creative Core(the direction predicted).

The model shown in full in Table 7 – CreativeProfessionals as the dependent variable with allforeign-born citizens as a dependent variable – hadthe greatest explanatory power. In total, this modelexplains around 70 per cent of the distribution ofthe Creative Professionals. All of the independentvariables are significant at the 99 per cent level,with the exception of Openness (95%).

The location of the Bohemians, Openness andthe Cultural Opportunity index is positively cor-related with the localization of the Creative Pro-

fessionals. This means that, as we would expectfrom the theory of the Creative Class, whereverthese quality of place indicators are high, levels ofCreative Professionals will also tend to be higher.Moreover, from the standardized coefficientsshown in Table 7 we can infer that of these varia-bles it is the presence of the bohemians that has thegreatest influence. With regard to the other two in-dependent variables, as expected unemploymentis negatively associated with the location of theCreative Professionals; the nature of causality be-hind this relationship is open to debate. On the onehand it may represent an association with higherlevels of social cohesion, while on the other itcould be seen as purely labour market related inthat creativity is a growing area of employmentand as such would be expected to coincide withlower unemployment. Finally, the public provi-sion index is negatively linked to the distributionof the Creative Professionals, which is the oppo-site of what might be expected. This could be dueto the nature of public sector employment in theUK, which tends to be proportionally higher inless prosperous areas, reflecting a lack of privatesector jobs in combination with a conscious policyof employment redistribution. On the one hand,Tables 6 and 7 would appear to be prima facie ev-idence of systematic variation between sections ofthe Creative Class; however, we need to be waryof data or other specification issues; this is a cleararea for further research and one in which quali-tative (casual) evidence will play a useful part inthe near future.

Table 7. Results of the Combined Model: detail.

Unstandardized Standardizedcoefficients coefficients Collinearity statitics

Model B Std. error Beta t Sig. Tolerance VIF

1 (Constant) 27.882 1.097 25.406 .000Openness .143 .060 .218 2.360 .020 .345 2.897Bohemians 3.235 .906 .414 3.572 .001 .221 4.530Public provision –.423 .126 –.223 –3.356 .001 .670 1.492Cultural opportunity .877 .265 .302 3.313 .001 .356 2.807Unemployment –.903 .145 –.385 –6.231 .000 .776 1.289

Note:Dependent variable: Creative Professionals

Model summary

Model R R square adjusted R square std. error of the estimate

1 .844(a) .713 .698 2.19272

NICK CLIFTON


The creative class and indicators of prosperity, growth and technologyHaving examined where the creative class is locat-ed in England and Wales, and how this distributionis associated with various indicators of quality ofplace, attention is now turned to the relationship be-tween the creative class and basic indicators ofprosperity. Constraints of space prevent our report-ing a full analysis of the geographical variationswithin each of these indicators here; thus Table 8shows overall bivariate correlations between thethree creative class groupings and the five indica-tors of prosperity, growth and technology.

The association between the two general indica-tors of prosperity (i.e. population growth and em-ployment growth) and the localization of the Cre-ative Class is statistically significant and positive.The relationship between Creative Class locationand population growth is quite strong for both theCreative Class as a whole and the Creative Profes-sionals, but weak for the Creative Core. For em-ployment growth, the relationship is however quiteweak for both the Creative Class and its two sub-groups. Overall therefore, a high concentration ofthe creative class tends to be found in places thathave growing populations and rising employment.This may be interpreted as evidence in support ofthe ‘jobs follow people’ aspect of the Florida thesis;that is, that the creative class creates prosperity ingeneral by its very presence. Bivariate correlationscan however only indicate association (i.e. covari-ance) and do not themselves imply any causal re-lationships between the variables involved. Putsimply, it may be that the creative class also followsprosperity, rather than creating it. Conclusions re-garding causality will need to be informed by morequalitative data, currently being gathered (at thetime of writing).

With regard to employment in the Tech-pole sec-

tors, the relationship between the Creative Classand share of “high-tech” employees is positive andsignificant; this relationship is somewhat strongerfor the Creative Class in general and the CreativeProfessionals in particular than it is for the CreativeCore. This means that a high concentration of res-ident creative class tends to be associated with thepresence of relatively high levels of employment intechnology-based businesses. This preponderancefor the creative class to be collocated with high-technology activities is consistent with what is ob-served in the North American research (Florida2002a, 2002b, 2002c; Gertler et al. 2002). There ishowever no evidence to suggest that the presence ofthe creative class (in any of its guises) is associatedwith any growth in employment in these activities,these correlations not being significantly differentfrom zero. This is on the face of it a major differ-ence between the UK and the US/Canada creativeclass results, and one that requires further investi-gation given the central tolerance and diversity be-gets creativity begets technology-based employ-ment growth theme of the Creative Class model.

Finally, the correlation between the presence ofthe creative class (in all its forms) and the rate atwhich new firms are created is positive and signif-icant. The relationship is strong with respect to theCreative Class as a whole, and for the Creative Pro-fessionals but less so for the Creative Core. Thismeans that localities in which the creative class inEngland and Wales is concentrated typically exhib-it higher levels of new firm formation; this is con-sistent with the Florida thesis but again open to in-terpretation with regard to causality.

Conclusions and issues for further researchOur first research question was concerned withmapping the size and distribution of the creative

Table 8. Indicators of prosperity – bivariate correlations.

Correlation with Correlation with Correlation withCreative Class Creative Core Creative Professionals

Population change 0.51** 0.29** 0.56**Employment change 0.27** 0.23** 0.25**Milken Tech-pole index 0.64** 0.49** 0.63**Employment change (hi-tech) 0.04 0.01 0.05New firm formation 0.76** 0.46** 0.82**

Source: Office of National Statistics, various datasets14

Notes:* Significant at the 95% level.** Significant at the 99% level.



class in England and Wales. In its widest defini-tion it currently accounts for just over 37 per centof all employment, yet the dispersion of the cre-ative class throughout England and Wales issomewhat uneven. High concentrations are typi-cally an urban and metropolitan phenomenon, butnot exclusively so, a number of inner city areashaving some of the lowest rankings observed, incommon with a number of localities and regionsafflicted by the long employment decline withinold traditional industries. It should be noted thatthe level of geography employed does have an im-pact here, and striking the balance between high-lighting distinct localities and functional labourmarkets on the one hand and being able to observeinteresting variations and idiosyncrasies on theother is not always easy. Particularly within thelarger cities, the findings of Markusen (2006) andWojan et al. (2007) suggest the need to investigatethe subtleties of bohemian and creative class dis-tribution at the sub-metropolitan (or even neigh-borhood) level.

Attention was focused upon how variations inquality of place are related to the location of the cre-ative class. This analysis of this second researchquestion showed statistically significant relation-ships between the localization of the creative classand these indicators. Tested individually, all the in-dicators were significantly correlated in the direc-tions hypothesized with the exception of the PublicProvision index, which was partially so. Thestrength and significance level of these correlationsshowed some variation by creative class type (i.e.Creative Class as a whole, Creative Core and Cre-ative Professionals). The combined quality of placeregression model produced high levels of explana-tory power with independent variables largely sig-nificant in the direction suggested by Florida’s the-ory: there are however some variations dependingupon which creative class subgroup is employed asthe dependent variable.

Do these variations reflect “real” differences in,for example, preferences for quality of place? Al-ternatively are they more likely to derive from data/model specification issues? These observed differ-ences between the Creative Class groupings under-line the potential value in seeking to unpack thesometimes amorphous set of occupations included;the knowledge base approach outlined by Hansenet al. (2005) may represent a fruitful method of do-ing just that. Essentially this requires the re-codingof the Creative Class dataset (or gathering of a newone) into (in their methodology) the users and cre-

ators of symbolic, analytical, and synthetic knowl-edge. A valid alternative (or indeed complement) tothis approach might be to conduct a series of par-allel focused studies on the distribution of narrowlydefined occupational groups; that is, to minimizethe conflation of identities, processes and identi-ties: the bohemians might represent the most ready-made starting point for such an exercise – averag-ing around 2 per cent of the labour force. Anotherway to achieve greater understanding could be toexplore life-stage effects of location choice, for ex-ample, as per the Markusen and Johnson (2006)study of artists in Minneapolis.

More generally it may also be insightful to bringtogether research on the hard and soft infrastruc-tures of a given place. For example, what types ofspaces, natural or built environments might aid (orhinder) milieu effects? At present these are essen-tially assumed to happen inside a spatial black box.

Indicators of prosperity, technology and growthare all significant and positively associated with thepresence of the creative class, and again variationsin the significance and strength of relationshipswere observed by creative class subgroups. The no-table exception to this pattern was technology-based employment growth, which showed no asso-ciation with the location of the creative class. Thissuggests the representation of both technology en-trepreneurs and bohemians in the same places inNorth America while in the UK, by contrast, bohe-mians are concentrated in the London area and ina few other creative centres (Brighton, for exam-ple); technology entrepreneurship is less commonbut may also be found in smaller university cities.

Overall, we generally conclude that the creativeclass in England and Wales displays similar prop-erties to those ascribed to it in the USA (Florida2002a, 2002b, 2002c) and Canada (Gertler et al.2002), albeit with certain caveats as noted. It shouldalso be noted here that although our quantitative re-sults provide evidence of consistency with the Cre-ative Class theory through numerous correlationsand associations, it is much more difficult to drawinferences relating to actual causality. This will bean issue that the qualitative part of the research ad-dresses, with qualitative interviews and case stud-ies designed to uncover the motivations andthought processes behind locational choices (i.e.designed to probe the causations that underpin theobserved correlations). A good example of this isthe association of creative class and openness: doesthis correlation between the creative class and di-versity arise merely from a more or less coinciden-

NICK CLIFTON


tal location in certain metropolitan and urban areaswith little or no interaction, is it primarily a kind ofconsumption relationship in which the creativeclass are able to access, for example, exotic ethnicshops and a wide variety of interesting restaurants,or is it a reflection of the creative class’ genuinepreference to live and work in an open, tolerant anddiverse community as Florida describes? Each ofthese outcomes would potentially be consistentwith our findings so far.

On the issue of diversity, Fig. 2 suggests anumber of places that have arrived at similar levels(as we are able to measure it) but almost certainlyfor very different reasons, and with different out-comes likely. Clearly the whole issue of diversityand tolerance needs further exploration; one possi-bility is using the census data to examine variationswithin the overall foreign-born or non-Western fig-ures; at present it is not possible to distinguish be-tween places that have equal-sized foreign-borncommunities originating from one or a hundred dif-ferent parts of the world: intuition alone would sug-gest that two such places are likely to be very dif-ferent. In addition, on the wider point of measuringtolerance, one of the key insights of Florida andGates (2001) is the linking of gay and hi-tech indi-ces, something not possible in the UK using sec-ondary data. Brighton is the classic example fromthe UK of a locality whose position in the tolerancerankings would be significantly affected if this al-ternative measure were used. To widen this pointout, Florida himself suggests that the USA is nowstarting to lose its dominance in what he refers toas ‘the new global competition for talent’ (Florida2004); if this is indeed the case, what might be thepotential impact on both policy and potential eco-nomic outcomes for Europe in general, and the UKmore specifically?

Perhaps the most pressing research prioritythough is the fate of the non-creative class, whichafter all comprises more than 60 per cent of the la-bour force. Will they be limited to those meagretrickle-down benefits of creativity, or can every-one become more creative; and who will be re-sponsible for making this happen? The urbanleaders driving forward the Creative Class agen-da? National governments? Or will it be left to thecreative class themselves to “grow up” and takeresponsibility as Richard Florida himself propos-es? Analogously, there are similar questions to beresolved regarding non-creative places. Peck(2005) warns against the creation of “cargocults”, vainly attempting to flag down footloose

creativity much in the same way that cities and re-gions used to enter into zero-sum games chasingold-style inward investment. Perhaps it would bean interesting exercise for further research to re-visit the sites of creativity attraction programmes,to systematically evaluate their impacts, success-ful or otherwise.

AcknowledgementsResearch was funded by the UK Economic and So-cial Research Council as a project entitled ‘Tech-nology, Talent and Tolerance in European Cities: AComparative Analysis’ (Grant No. RES-000-23-0467), and we extend our gratitude to all the Euro-pean partners for their various efforts in ensuringthe project was interesting and enjoyable for allconcerned. The author is grateful to the sponsorsfor making this research possible. The author isalso thankful for the insightful comments providedby the Editor and anonymous referees. The usualdisclaimer applies.

Notes1. At present data from Northern Ireland and Scotland are ab-

sent from our analysis. This omission will be addressed dur-ing subsequent iterations of the analysis.

2. Markusen (2006) proposes that Florida’s results in the USAactually derive at least in part from human capital effects,rather than from creativity effects per se. However, short ofassessing all work-related creativity on an individual basis,proxy by occupation is the only way large-scale studies ofthis nature can be undertaken realistically.

3. Peck (2005) also notes the use of disarming terminologythat invites positive association, and asks who indeed wouldseek to be rigid, narrow-minded and conservative when theycould be flexible, open and creative?

4. Indeed, as Markusen (2006) herself notes, visual artists may(for example) have some different locational preferences to,say, performing artists.

5. There has been some work on Europe (Florida and Tinagli2004) but this been at the national level only. Within this,with regard to creativity, the traditional large economies (in-cluding the UK) are seen as losing ground, Southern Europecontinues to lag behind, while the Netherlands and the Nor-dics forge ahead.

6. This involves the construction of a dataset that allows threedistinct fields of analysis: examining how the Creative Classmodel applies within each partner country, the performingof inter-country comparisons across partner countries, andfinally a comparison between the “TTT group” as a wholeand existing US and Canadian studies. Results relating tothese latter two project aims will be reported elsewhere indue course.

7. As we define it, the Creative Core maps very closely to themajor occupational group Professional Occupations, albeitwith a small number of occupations therein placed in ourCreative Professionals category.

8. Mainland European countries use the International Stand-



ard Classification of Occupations (ISCO) system; due tothe nature of the wider research being undertaken it waswith reference to this system that our occupational group-ings (bohemian, creative class) were defined. In many cas-es ISCO maps directly to SOC2000, but in other cases itdoes not.

9. I.e. those people who have not worked more than one hourduring the short reference period regardless of whether ornot they are in receipt of unemployment related benefits.

10. Not in terms of absolute size – the USA and Canadian re-gions typically being much larger – but rather having a sim-ilar role within the national context in question.

11. This figure rises slightly to 0.28 for the bohemians only.12. This statistic should not obscure the fact that massive vari-

ation exists within London, with some very low CreativeCore LQs found therein (for example, Barking and Dagen-ham, at 0.51 the third lowest in England and Wales).

13. It is worth noting here that although not a directly compa-rable analysis, there are generally some consistencies withthe indices produced by the Demos think tank (Demos2003).

Nick CliftonCentre for Advanced StudiesCardiff University44–45 Park PlaceCardiff CF10 3BBWalesUnited KingdomE-mail: [email protected]

ReferencesBRADFORD, N. (2004): Creative cities: structured policy dia-

logue report. Background paper F45, Family Network, Cana-dian Policy Research Networks, Ottawa, ON.

BROOKS, D. (2000): Bobos in Paradise: The New Upper Classand How They Got There. Simon & Schuster, New York.

BROWN, P. and LAUDER, H. (2006): ‘Globalisation, knowl-edge, and the myth of the magnet economy’, Globalisation,Societies and Education 4 (1): 25–57.

CANNON, T., NATHAN, M. and WESTWOOD, A. (2003):Welcome to the Ideopolis. The Work Foundation, London.

COOKE, P. and DE LAURENTIS, C. (2002): The index of knowl-edge economies in the European Union: performance rank-ings of cities and regions. Regional Industrial Research Re-port 41, Centre for Advanced Studies, Cardiff University,Cardiff.

CUSHING, R. (2001): Creative capital, diversity and urbangrowth. Unpublished paper, University of Texas at Austin,Austin, TX.

DCMS (2004): Culture at the Heart of Regeneration. Departmentfor Culture Media and Sport, London.

DEMOS (2003): ‘Manchester is favourite with “new bohemi-ans”’ [WWW document]. URL http://www.demos.co.uk/media/pressreleases/bohobritain [accessed in November2007].

DEVOL, R., BEDROUSSIAN, A. and KIM, S. (2007): Best Per-forming Cities 2007: Where America’s Jobs are Created andSustained. Milken Institute, Santa Monica, CA.

FLORIDA, R. (2000): Competing in the Age of Talent: Quality ofPlace and the New Economy. R.K. Mellon Foundation, Pitts-burgh, PA.

FLORIDA, R. (2002a): ‘Bohemia and economic geography’,Journal of Economic Geography 2 (1): 55–71.

FLORIDA, R. (2002b): The Rise of the Creative Class. BasicBooks, New York.

FLORIDA, R. (2002c): ‘The economic geography of talent’, An-nals of the Association of American Geographers 92 (4): 743–755.

FLORIDA, R. (2004): The Flight of the Creative Class: The NewGlobal Competition for Talent. HarperCollins, New York.

FLORIDA, R. and GATES, G. (2001): Technology and toler-ance: the importance of diversity to high-technology growth.Survey Series 1, Center on Urban and Metropolitan Policy,The Brookings Institution, Washington, DC.

FLORIDA, R. and TINAGLI, I. (2004): Europe in the CreativeAge. Carnegie Mellon Software Industry Center and Demos,Pittsburgh, London.

GERTLER, M.S., FLORIDA, R., GATES, G. and VINODRAI,T. (2002): Competing on creativity: placing Ontario’s citiesin North American context. A report prepared for the OntarioMinistry of Enterprise, Opportunity and Innovation and theInstitute for Competitiveness and Prosperity. Munk Centrefor International Studies, University of Toronto, Toronto,ON.

GLAESER, E.L. (1998): ‘Are cities dying?’, Journal of Econom-ic Perspectives 12 (2): 139–160.

HALL, P. and SOSKICE, D. (2001): ‘An introduction to varie-ties of capitalism’, in HALL, P. and SOSKICE, D. (eds): Va-rieties of Capitalism: The Institutional Foundations of Com-parative Advantage. Oxford University Press, Oxford, pp.1–68.

HALL, P., BREHENY, M., MCQUAID, R. and HART, D.(1987): Western Sunrise: The Genesis and Growth of Brit-ain’s Major High Tech Corridor. Allen & Unwin, HemelHempstead.

HANSEN, H.K., VANG, J. and ASHEIM, B.T. (2005): ‘The cre-ative class and regional growth: towards a knowledge basedapproach’, paper presented at the Regional Growth AgendasConference, Regional Studies Association, Aalborg, 28–31May.

HEATH, J. and POTTER A. (2006): The Rebel Sell: How theCounter Culture Became the Consumer Culture. Capstone,Chichester.

JACOBS, J. (1961): The Death and Life of Great American Cities.Random House, New York.

JACOBS, J. (1969): The Economy of Cities. Random House, NewYork.

LEITNER, H. (1990): ‘Cities in pursuit of economic growth: thelocal state as entrepreneur’, Political Geography Quarterly 9(2): 146–170.

LUCAS, R.E. (1988): ‘On the mechanics of economic develop-ment’, Journal of Monetary Economics 22 (1): 3–42.

MARKUSEN, A. (2006): ‘Urban development and the politics ofa creative class: evidence from a study of artists’, Environ-ment and Planning A 38 (10): 1921–1940.

MARKUSEN, A. and JOHNSON, A. (2006): Artists’ Centers:Evolution and Impact on Careers, Neighborhoods and Eco-nomies. Hubert H. Humphrey Institute of Public Affairs, Uni-versity of Minnesota, Minneapolis, MN.

MARSHALL, A. (1920): Industry and Trade. Macmillan, Lon-don.

PARR, J.B. (2005): ‘Perspectives on the city-region’, RegionalStudies 39 (5): 555–566.

PECK, J. (2005): ‘Struggling with the creative class’, Internation-al Journal of Urban and Regional Research 29 (4): 740–770.

PORTER, M.E. (2000): ‘Location, clusters, and company strate-gy’, in CLARK, G.L., FELDMAN, M.P. and Gertler, M.S.

NICK CLIFTON


(eds): The Oxford Handbook of Economic Geography. Ox-ford: Oxford University Press, pp. 253–274.

THRIFT, N.J. (2001): ‘“It’s the romance, not the finance, thatmakes the business worth pursuing”: disclosing a new marketculture’, Economy and Society 30 (4): 412–432.

WOJAN, T.R., LAMBERT, D.M. and MCGRANAHAN, D.A.

(2007): ‘Emoting with their feet: Bohemian attraction to cre-ative milieu’, Journal of Economic Geography 7 (6): 711–736.

WONG, C. (2001): ‘The relationship between quality of life andlocal economic development: an empirical study of local au-thority areas in England’, Cities 18 (1): 25–32.

Documents

THE “CREATIVE CLASS” IN THE UK: AN INITIAL ANALYSISnaumannj/professional geography... · THE “CREATIVE CLASS” IN THE UK: AN INITIAL ANALYSIS by Nick Clifton CLIFTON, N. (2008):