Creative knowledge workers and location in Europe and North America a comparative review CIJ 2(1)...

CIJ 2 (1) pp. 73–89 © Intellect Ltd 2009 73

Creative Industries Journal Volume 2 Number 1 © 2009 Intellect Ltd

Article. English language. doi: 10.1386/cij.2.1.73/1

Keywords creativity knowledge economy location economic outcomes

Creative knowledge workers and location in Europe and North America: a comparative review

Nick Clifton University of Wales Institute, Cardiff

Phil Cooke Cardiff University

Abstract Much of the recent interest in the development of individual creativity has drawn upon Richard Florida’s (2002a) book The Rise of the Creative Class. Whereas in the industrial age, classical and neoclassical economic theory told us that ‘people followed jobs’, in the modern knowledge economy Florida describes how ‘jobs fol-low talented people’. The research reported in this article represents an analysis of quality of place and the dispersion of the creative knowledge workers in seven European countries and builds upon the work that has been undertaken in North American cities in order to understand whether similar processes concerning the relationship between creativity, human capital, and high-technology industries are at work in Europe as claimed is the case within North America. Economic outcomes from creative class location are also reviewed. Finally, we consider the implications for further research, given the evidence presented here suggesting some variation of results by national socio-economic context.

IntroductionThe creative industries and occupations are an important source of employ-ment growth. According to a recent NESTA study, creative employment now accounts for approximately 1.9 million people or 7 per cent of employment in the UK alone, up from 0.9 million people or 4 per cent in the early 1980s (Higgs, Cunningham and Bakhshi 2008). Creative industries are often posi-tioned at the centre of many urban and regional development policies in the UK, becoming a delivery vehicle for economic growth through accumulation of human capital working in the industries. This trend is highlighted by a recent report by the Department of Culture, Media and Sport which states ‘Culture drives regeneration in many ways, from inspiring landmark buildings through to reviving the decaying centres of market towns to bringing a community together around an arts event’ (DCMS 2004).

Definitions of the ‘creative industries’, as exemplified by the DCMS in the UK, have typically been sectoral in approach (for example see http://www.culture.gov.uk/what_we_do/Creative_industries/ for a full listing of the industries included). While this type of methodology is not without merit, we feel that an approach based around the analysis of occupational data offers greater insight when examining the locational preferences of

CIJ_2.1_art_Cooke_073-090 73CIJ_2.1_art_Cooke_073-090 73 12/10/09 10:06:02 AM12/10/09 10:06:02 AM

74 Nick Clifton and Phil Cooke

creative and knowledge workers; not least as such a typology captures those creative individuals embedded in other industries not defined as ‘creative’ (Higgs et al. 2008).

Such accumulation of human capital is imperative to a region’s suc-cess in the knowledge economy as the value creation in many sectors thereof rests, increasingly, on non-tangible assets. The book value of intan-gible assets compared to raw materials has shifted from 20:80 in the 1950s to 70:30 in the 1990s (De Laurentis 2006; Cooke, De Laurentis, Tödtling and Trippl 2007). Places with a greater stock of highly educated people are, in general, more innovative and grow more rapidly and robustly over time.

Creative knowledge workers – a ‘creative class’Much of the recent interest in the development of individual creativity has drawn upon Richard Florida’s (2002a) book The Rise of the Creative Class. Whereas in the industrial age, classical and neoclassical economic theory told us that ‘people followed jobs’, in the modern knowledge economy Florida describes how ‘jobs follow talented people’. That is, places that dis-play ‘creative class’ characteristics, meaning a high presence of profession-als, technologists and bohemians, have performed best economically in recent years.

In a knowledge-based economy, the ability to attract and retain highly skilled labour is therefore perceived as crucial to the current and future prosperity of regions as well as entire nations. For example, Florida (2000) has argued that regions develop advantage based on their ability to quickly mobilize the best people, resources and capabilities required to turn inno-vations into new business ideas and commercial products. In particular, the ability to attract creative people in arts and culture fields and to be open to diverse groups of people of different ethnic, racial and lifestyle groups provides distinct advantages to regions in generating innovations, growing and attracting high-technology industries, and spurring economic growth (Gertler, Florida, Gates and Vinodrai 2002).

This research demonstrates that quality of place must be understood in broader terms than we have traditionally been accustomed to: while the attractiveness and condition of the natural environment and built form are certainly important, so too is the presence of a rich cultural scene and a high concentration of people working in cultural occupations (most specifi-cally the ‘bohemians’). According to the results from Florida and his associ-ates’ research, the underlying hypothesis is that the presence and concentration of bohemians in an area creates an environment of milieu that attracts other types of talented or high human-capital individuals. The presence of such human capital in turn attracts and generates innovative, technology-based industries (Florida 2002a).

However, given the interest Florida’s writings have received from aca-demics, policymakers, and the media alike, it is no surprise that they have been the object of a high degree of critical examination. This critique has centred most notably around the apparent fuzziness of some of the con-cepts, definitions and causal logic Florida employs, the seemingly conven-ient appeal of his ideas to the agendas of a multitude of urban actors and policymakers, and, conversely, the minimal attention paid to difficult issues


75Creative knowledge workers and location in Europe and North America

1. There has been some work on Europe (Florida and Tinagli 2004) but this has been at the national level only. Within this, with regard to creativity, the tradi-tional large economies (including the UK) are seen as losing ground, southern Europe con-tinues to lag behind, while the Netherlands and the Nordics forge ahead.

2. At present Northern Ireland and Scotland are absent from our analysis, due to data availability issues. It is intended that this will be addressed during subsequent iterations of the analysis.

3. Deriving from the ‘3 Ts’ of Technology, Talent and Tolerance.

such as the potential inequalities and negative externalities implied by a cre-ative class model of regional development (see for example Peck 2005). In more recent work however, Florida does consider at some length the ever-increasing levels of economic polarization that have typically been the flip side of economic development in recent years: indeed, he now regards the creative class as a ‘privileged minority’, who connect with each other from across the global peaks of an increasingly ‘spiky world’ (Florida 2007: 33).

Research issuesDespite commonalities, there are aspects of the respective socio-political structures of North American and European economies that show strong divergence, for example with regard to political priorities, functioning of labour markets, economic growth processes and social outcomes (for example, Hall and Soskice 2001). Given that virtually all of the systematic research of the phenomenon has taken place in North America, the ‘crea-tive class’ warrants investigation in a pan-European context.1 Therefore, the research reported in this article represents an analysis of quality of place and the dispersion of the creative class in seven European coun-tries, building upon the work described above that has been undertaken in North American cities in order to understand whether similar proc-esses concerning the relationship between creativity, human capital, and high-technology industries are at work in Europe as is claimed is the case within North America.

The research took place within the context of a project entitled ‘Technology, talent and tolerance in European cities: a comparative analy-sis’, involving matched data sets and research partners in Denmark, Sweden, Norway, Finland, the Netherlands, Germany, and the UK.2 This involved the construction of a data set that allowed three distinct fields of analysis; examining how the creative class model applies within each part-ner country, the performing of inter-country comparisons across partner countries, and finally a comparison between the ‘TTT group’3 as a whole and existing US and Canadian studies. This article then seeks to:

• Briefly review the distribution of the creative class within the seven European research partners

• Examine the impact of quality of place upon this dispersion• Consider the connection between the location of the creative class and

inequalities in technical and economic outcomes• Introduce some comparative data from the North American creative

class research

To explore these issues, the role of human capital, creative capital, and diver-sity in technology-based economic development is investigated. The research uses the two measures developed by the North American studies: the ‘crea-tivity (bohemian) index’ to reflect creative capital, and the ‘diversity (mosaic) index’ to reflect openness and diversity. This suggests that there will be a relationship between openness to creativity and diversity, and the ability to support high-tech industries and economic development based on talented workers. New indices are developed in order to grasp the fundamental dif-ferences in certain aspects of economic life between North American and



4. The Office for National Statistics in the UK, Statistics Sweden, Statistics Denmark, Statistics Norway, Statistics Finland, Statistics Netherlands, and Statistiches Bundesamt Deutschland.

5. As we define it, the creative core maps very closely to the major occupational group ‘professional occupations’ in the UK, albeit with a small number of occupa-tions therein placed in our ‘creative profes-sionals’ category.

European societies (see the following section). As noted, this kind of analy-sis has not yet been performed for European cities, and has the potential to shed important new light on the role of quality of place in shaping the com-petitiveness of city regions in Europe.

Data and methodologyThe data set for the quantitative analyses was constructed using data sup-plied by the official statistics agencies of the various European partners.4 In terms of the individual data sources involved in this, these obviously vary from agency to agency but are analogous to the UK’s ‘census of popula-tion’, supplemented by the Labour Force Survey and the Annual Business Inquiry. As such, only the specific UK data source is listed for each individ-ual variable described below. In general, the variables used were designed to jointly maximize consistency between the different European countries involved in the wider project, and between Europe and the US and Canada. Variation in availability across the partner nations has inevitably imposed constraints upon the data used by individual partners, and in some cases involved compromises around ‘lowest common denominator’ levels of detail and time frames. Various indices are constructed from this data as described below, and explored through means of plots, regressions and correlations.

Key variables for quantitative analysesThe key variables for the quantitative analyses are the ‘bohemian index’, the ‘talent index’, the ‘diversity index’, and the ‘tech-pole index’. These mirror variables employed in previous research by Florida (2002a, 2002b, 2002c) and Gertler, Florida, Gates and Vinodrai (2002) on the geography of talent and the rise of the creative class. In addition indicators for cultural and rec-reational amenities, which were also used in Florida’s studies, will be consid-ered. A pair of new indicators – developed to reflect the characteristics of European cities and their national political economies – is also introduced. These two new indicators are social cohesion, and a ‘public provision index’ measuring the supply of public sector goods, for example education, health care and social security.

Creative classClearly of central importance is the ability to actually quantify the size of the creative class present in any given location. These are essentially peo-ple who are involved in the creation of new knowledge as a key constitu-ent of their work, or use existing knowledge in new ways, combinations and so on. In the absence of a primary data set relating to the actual engagement in such activities, this is approximated by the use of occupa-tional categories. We subdivide the creative class into the ‘creative core’ (scientists and engineers, architects and designers, academics and teach-ing professionals),5 and the ‘creative professionals’ (associated profes-sional and technical occupations of the creative core, managers, financial and legal professionals). Data for this index (and the others using occu-pational data) were derived from the 2001 ‘census of population’, the only source with sufficient size to allow four-digit standard occupational



6. Mainland European countries use the ‘international stand-ard classification of occupations’ (ISCO) system; due to the nature of the wider research being undertaken it was with reference to this system that our occu-pational groupings (bohemian, creative class) were defined. In many cases ISCO maps directly to SOC, but in other cases it does not, requiring the manual adjust-ment of UK categories using detailed four-digit data.

classification (SOC2000) breakdowns at the levels of regional disaggrega-tion required.

Bohemian indexThe ‘bohemian index’ is defined using employment in artistic and creative occupations. It is a locational quotient that compares the region’s share of the nation’s ‘bohemians’ to the region’s share of the nation’s population (local prominence of employment in artistic/creative occupations com-pared to national prominence of employment in the same occupations). Data for the UK is delineated using the SOC2000 system; within this it is convenient to use sub-major group 34 (culture, media and sports) to define the bohemian occupations.6

Cultural opportunity indexThe opportunity to enjoy cultural and recreational activities can be calcu-lated more directly than other indicators on quality of place. According to Florida, such opportunities play an important role in cities’ ability to attract the creative class. The proportion of employees in the cultural and recrea-tional industries within an area is used as an indicator of this cultural opportunity. This involves using a number of three-digit standard industrial classification (SIC) groups from the Annual Business Inquiry to account for employment within restaurants and bars, libraries, museums and other cul-tural and entertainment activities.

Talent indexTalent is defined as the proportion of the population over 18 years of age with a bachelor’s degree or higher (local proportion compared to national). This data was obtained from the 2001 ‘census of population’.

Diversity (openness) index The ‘diversity index’ is the European counterpart of Florida’s ‘melting pot index’. It is calculated as the proportion of the total population that is for-eign-born. A second index, which measured the proportion of total popula-tion that is foreign-born of non-western origin, was also constructed; the logic behind this was that a more visible manifestation of diversity may be a more valid tool, or at least provide greater insight into the concept of tol-erance; this data was provided by the 2001 ‘census of population’.

Tech-pole indexFollowing the method of the Milken Institute (e.g., DeVol, Bedroussian and Kim 2007), this index shows local employment in technology-intensive man-ufacturing and service sectors (specialization and size). The index compares a region’s share of national employment in high-technology industries to the region’s overall share of national employment; this is then adjusted by multi-plying by a region’s overall share of national high-technology employment. Therefore, it reflects both the region’s degree of specialization in technology-intensive activity as well as its sheer scale of employment in these sectors. The index includes technology-intensive sectors in both manufacturing and services. Data for constructing the index was extracted from the Annual Business Inquiry (ABI) employee analysis for 2002.



7. i.e. those people who have not worked more than one hour during the short reference period regardless of whether or not they are in receipt of unemployment related benefits.

8. Not in terms of absolute size – the US and Canadian regions typically being much larger – but rather having a similar role within the national context in question.

Social cohesion indexFlorida himself has expressed concern regarding the polarization in terms of prosperity that high levels of creative class employment has been associ-ated with in the US – to the point where this is becoming a negative quality of place factor for the creative workers themselves. As such it was felt that some attempt to factor this in to the research design should be made; i.e., not cohesion in the sense of a social capital indicator (which may be incom-patible with the concept of openness), rather one based on perceived ine-quality. Ideally this should have involved some kind of ‘Gini coefficient’ with respect to income distribution, but this kind of data was not available at the spatial level we are using (this was also true for the other European project partners). As such, the apparent level of exclusion from mainstream eco-nomic activity was used, as represented by the International Labour Organization (ILO) definition of unemployment.7 Using a labour market variable as a proxy for social cohesion introduces a new net of complicating factors – as with other issues relating to operationalisation of creative class concepts, these are revisited in later sections. The data was taken from the Labour Force Survey with respect to the twelve-month period March 2001 to February 2002.

Public provision indexHaving acknowledged the potential implications for social cohesion implied within the creative class model, there is a need to further take account of the European context; given the high levels of private employment-based provision in the USA, particularly for health care, to what extent might high-quality social welfare provision be a quality of place factor for the European creative class? In this index, levels of employment in education and health care (SIC2003, two-digit codes 80 and 85) for a given locality are expressed as a proportion of the resident population, using data from the Annual Business Inquiry.

Levels of GeographyAs described above, data was collected so that the triple goals of country-specific, inter-country and Europe-North America comparisons could best be achieved. This was also true with respect to the levels of geography i.e., spatial units employed in the analysis. With the partner countries account-ing for large variations, for example in size, governance structure, patterns of population dispersion, it was impractical to impose a single standard definition; in practice such definitions actually have different meanings dependent upon the context in which they are applied; i.e., this definition would actually define a spatial unit that has a different meaning from one country to another.

As we were seeking definitions that encapsulated something approach-ing functional labour markets (analogous at least in part to the munici-pal city regions employed in the North American research8) it was decided that the ‘most meaningful functional unit’ in each national context would be used – subject to this also being a level at which the necessary statis-tical data was available from the relevant national agencies. As Parr (2005) notes, standard UK administrative geographies do not typically relate in a systematic way to any theoretical construct of the city region;



for the UK this meaningful unit was primarily the ‘Nomenclature of Territorial Units for Statistics’ (NUTS) level three definition (105 spatial units in England and Wales). Due to the sometimes complex and non-hierarchical nature of these standard geographies, this is supplemented by analysis using the ‘unitary/county authority level’ (171 units). With regard to the levels of geography employed by the other European partners, this essentially broke down into two distinct groups; the Scandinavian countries typically based their data on spatial units at the level below NUTS three (usually data relating to municipalities) and then partially recombined these into a smaller number of units, to construct (for exam-ple) city regions around large metropolises (i.e., capital cities) and, con-versely, sparsely populated larger rural areas. Data for Germany and the Netherlands was more similar to that of the UK in construction, i.e., based around the NUTS level three.9

A significant difference between our geography and that of the research undertaken in the US and Canada is that it is totally inclusive (i.e., no local-ities are excluded), whereas the North American analysis has tended to focus exclusively on large urban centres. This was essentially done for rea-sons of consistency across the project partners; the European countries are generally more densely populated than North America, the demarca-tion of city regions less obvious and so partial selection is potentially more arbitrary.

Results and discussionLocation of the creative classFigure 1a shows how the distribution of the creative class varies; the Scandinavian economies show quite consistent ranges of variation, slightly below that of the UK. The Netherlands produces the highest standard devi-ation, while at the other extreme the German figure is substantially lower than the others. There is then partial, but not complete, support for the idea that creative class will be less evenly distributed in what Hall and Soskice (2001) term ‘liberal market economies’, rather than their more coordinated counterparts.

With regard to absolute levels of creative class, the Netherlands has substantially higher membership than the other countries at well over 50 per cent of the workforce. The remainder are at quite similar levels, with the UK being the next highest. Conversely, creative class membership in Norway is significantly lower than the rest of our sample.

9. The actual numbers of units involved is as follows: Denmark 34, Finland 25, Norway 77, Sweden 70, Netherlands 40, Germany 438; as noted above, for the UK we used the 105 NUTS level three areas of England and Wales, supplemented by unitary/county data as appropriate.

UK Germany Norway Denmark Finland Sweden Netherlands

Range (%) 24.1 – 64.9 26.4 – 46.5 9.8 – 38.0 20.2 – 47.7 24.0 – 46.1 21.9 – 47.9 40.2 – 66.6

Mean (%) 36.3 33.3 18.6 27.6 33.4 29.8 52.4

Median (%) 35.1 32.8 17.9 26.8 32.4 28.6 50.6

Standard Dev.

6.1 3.6 5.4 5.8 5.1 5.3 6.9

Figure 1a: Comparative distribution of the creative class with bohemians.



Figure 1b shows the same analysis as 1a, but this time with regard to the creative core only. The most striking observation here is that for this the UK is now at the lower end of the range for unevenness of dispersion but the Netherlands remains relatively high. This result is significant in that it con-firms the possibility that differing national socio-economic systems may indeed have differential implications for the localization of the creative class, depending upon which specific element of the group is under consid-eration.

There may be other (but related) reasons that go some way towards explaining the variations between Figures 1a and 1b; for example there is some evidence to suggest that the creative core is even more of an urban phenomena that the wider creative class. This could mean that countries offering greater scope for the existence of regionally dispersed centres of creativity will demonstrate a more even distribution of the creative core. This is an issue beyond the scope of this initial comparative review, but one which merits further exploration.

Impact of quality of placeAs outlined earlier, Florida suggests that the creative class is highly mobile, with strong preferences for certain aspects of quality of place. In conjunction with Canadian researchers (see for example Gertler, Florida, Gates and Vinodrai 2002) he shows that North American cities with high levels of creative class tend to be open, tolerant and diverse places, with high levels of recreational and cultural opportunity. Thus, as described earlier, we have generated data on the following variables hypothesized to impact upon quality of place (and therefore locational preference) for creative knowledge workers – bohemian index, cultural opportunity index, diversity (openness) index, social cohesion, ‘public provision index’ (PPI). Restrictions of space preclude a full bivariate analysis here (i.e., discussion on a variable by variable basis of association with the location of the creative class – for this see Clifton 2008, with respect to the UK). However, given its inclusion in the creative class model specifically to reflect the European context, it is apposite to review the PPI briefly below.

Figure 2 shows how the public provision index varies across the seven nations in question. Absolute (mean) values are typically higher in the Scandinavian countries (particularly Norway and Sweden), which is consist-ent with what might be expected. Conversely, that of the UK is somewhat lower, while Germany and the Netherlands are substantially so. This in an interesting observation in itself, in that it suggests that ‘liberal’ markets and ‘coordinated’ markets (after Hall and Soskice 2001) do not necessarily


Range (%) 5.0 – 13.7 4.7 – 13.7 0.8 – 11.4 6.4 – 13.9 7.9 – 17.2 5.1 – 19.6 6.7 – 18.1

Mean (%) 9.3 8.1 3.7 9.0 11.6 8.9 12.2

Median (%) 9.1 8.0 3.1 8.7 10.9 8.3 11.5

Standard Dev. 1.8 1.9 2.4 1.7 2.6 2.3 2.4

Figure 1b: Comparative distribution of the creative core.



divide neatly into categories of ‘low’ and ‘high’ public provision, at least not as we define it here. There is, however, some support for the liberal versus coordinated taxonomy in that the standard deviation of the PPI is greatest with respect to the UK. Conversely, particularly low deviations are observed for Germany and Denmark. It is possible that an interaction between social policy and the particular physical geographies of Scandinavia may play a role here; in the Nordic countries the PPI tends to be low in regions with large populations – this is likely to be due to some kind of economies of scale effect. Large municipalities can, for example, accommodate more pupils into classes, and the civil servants can quite easily deal with a few more cases without significantly reducing quality of provision. In other words, the welfare system guarantees a minimum of welfare, and this mini-mum can be achieved despite fewer public employees per head in large metropolitan regions. For Germany, welfare provision is even and rural areas are predominantly well provided for, so the interpretation is a slightly different one.

In order to test a unified creative class model, the multiple regression method was employed. The advantage of using this method is that all indicators are tested in one model, and therefore it is possible to control any multi-collinearity between the independent variables. With regard to this multivariate analysis, Figure 3 shows the results of the regression models generated across the European research partners (the pre-existing North American results are not usable in this context here, due to lack of consistency with the parallel data collected in Europe). The results are quite varied but some points emerge consistently: the bohemian index, openness and the public provision index are the most consistently signifi-cant factors; these variables are significant with the signs hypothesized in the creative class model of locational attractiveness, with the exception of the public provision index. The bohemian variable is significant in four of the six countries included in the model, the exceptions being Norway and Denmark. The pattern for openness is slightly more complicated; positive and significant results are obtained for the UK, Germany and Norway, whereas for the Netherlands this variable is negative at the 99 per cent level of significance. This means that while the creative class is more evenly distributed in, for example Norway and Denmark, than in the UK or Finland, bohemians are more concentrated in the capital city. Moreover, given that in Finland and Sweden in particular, rural populations are often sparse, places with intense provision (towns and cities) are where a mod-erately evenly spread creative class tends to locate.


Range (%) 5.4 – 11.6 3.2 – 7.1 9.7 – 20.1 8.9 – 14.4 6.6 – 12.0 9.5 – 17.4 3.4 – 12.7

Mean (%) 9.3 4.9 12.6 11.9 9.4 12.6 5.3

Median (%) 8.9 4.9 12.5 11.8 9.5 12.5 4.8

Standard Dev. 2.1 0.8 1.8 1.1 1.6 1.4 1.8

Figure 2: Comparative variation within public provision index.



The results for the cultural opportunity index (the proportion of employees in the cultural and recreational industries) are interesting and would merit further investigation in new research; this variable is only sig-nificant in one of the six country-specific models – Germany. This may be a reflection of the varying geographies of culture itself, or it may be influ-enced by measurement issues – for example casual empiricism suggests that different types of bar or restaurant (recall the definition of this meas-ure discussed earlier) will play very different ‘cultural’ roles; the employ-ment data does not itself allow these distinctions to be made. Unemployment is significant in two of the six country models (UK and Germany), and in different directions; this perhaps reflects the ambiguity over what this variable is actually measuring – on one hand it serves as an (inverse) indicator of social cohesion, which the creative class are sup-posedly attracted to (or at least deterred by its absence), while on the other it acts as a labour market variable which one might expect to be negatively correlated with higher levels of creative class employment in a buoyant local or regional economy.

The public provision index results are also interesting in that they are positive and significant, with the exception of the UK and Netherlands. In a hypothetical situation where public provision is spread evenly and is of simi-lar quality this would make it irrelevant for any locational choices (regardless of preferences); this is unlikely to be the explanation for the observed pattern of results – the non-significant result for the UK may be linked to a greater emphasis on private provision for these services, or perhaps some other underlying reason for a difference in creative class preferences in this area.

With regard to the R2 values of the various regression models, there is some support for the idea that the creative class model tends to work more effectively in a liberal market rather than a coordinated market (Hall and Soskice 2001; i.e., essentially the UK in this research), but this is not exactly conclusive, with the value for Finland being the highest and the German model very similar to the UK.

UK Germany Norway Denmark Finland Netherlands

Bohemian .562 .459 (.006) (–.096) .448 .489

Openness (non-western) .197 .372 .496 (.172) (–.109) –.409

Public provision index (–.056) .460 .217 .507 .421 (.015)

Cultural opportunity index

(.087) –.235 (.162) (.150) (.258) (–.148)

Social cohesion –.415 .149 (–.086) (.254) (–.183) (–.025)

Adjusted R2 .664 .620 .398 .228 .828 .153

( ) = not significantDependent variable is European creative class location quotient (LQ), excluding Bohemians. Sweden excluded due to lack of regional unemployment data.

Figure 3: Results of regression models: regression coefficients.



The creative class and indicators of prosperity and growth Having reviewed the distribution of creative knowledge workers in the seven European partner nations, and how this is associated with various indicators of quality of place, attention is now turned to the relationship between the creative class and basic indicators of prosperity, i.e., the outcome variables.

Figure 4 shows the association between the relative concentration of the creative class, and levels of unemployment and employment change. In general we would expect to observe greater mobility of labour and more rapid structural adjustment within national economic systems that were more liberal in nature (i.e., more closely resembling those in the US), and less coordinated or corporatist (as typified by the Nordic countries). Thus, we would expect that the presence of the creative class will be more strongly associated with strong employment growth in the former that in the latter. The results shown in Figure 4 do not allow us to support this hypothesis – positive and significant correlations are observed for four of the six coun-tries for which data is available, and of these the value for Sweden is similar to that of the UK while the Finnish one is the highest. By analogous reason-ing it would be expected that the location of the creative class will be more likely to be significantly (and negatively) associated with unemployment in deregulated market regimes; from the figure we can see that results sup-port this hypothesis unequivocally – the observation for the UK being the only significant one obtained. This means that concentrations of creative class location and of unemployment are more polarized within the UK. An alternative explanation could be differing preferences: i.e., in other coun-tries there is no (or less) creative class preference against unemployment or perceived social injustice than in the UK. Such qualitative observations are beyond the scope of this article, but, albeit anecdotally, seem unlikely.

Finally, Figure 5 shows the correlation between the concentration of the creative class in a given locality and the observed rate of business start-ups; where data is available this is observed to be consistently positive and sig-nificant. Given the correlation observed for the UK, but also for Finland,


Employment change (93–02)

.526** –.041 .375** – .678** .559** .127

Unemployment –.255** –.008 –.115 .250 –.373 – –.008

* significant at the 95 per cent level.** significant at the 99 per cent level.

Figure 4: Correlation between creative class location, employment change and unemployment.


New firm formation .762** .586** .331** – .780** .316** –

* significant at the 95 per cent level.** significant at the 99 per cent level.

Figure 5: Correlation between creative class location and new firm formation rate.



there is partial support for the a priori expectation that there will be a stronger association between the localization of the creative class and lev-els of new firm formation in less coordinated economies.

Comparative results – Europe and North AmericaAs described earlier, the European ‘TTT group’ collected parallel data, intended to mirror preceding research undertaken in Canada and the US. The nature of the two data sets does not, however, allow them both be included simultaneously in unified national-level statistical models (the US creative class data for example includes the large municipalities only – i.e., it is not exhaustive for the whole county, unlike the TTT data). It is, however, possible to carry out some comparative analyses of matched individual vari-ables (or matched bivariates). Examples of this are presented briefly below.

As a selected example of the quality of place indicators employed in the research, Figure 6 shows cities in Europe and North America with the high-est concentration of bohemians. Amsterdam and London are significant outliers, with over twenty bohemians per 1000 population – nearly double the third placed city (Helsinki). Vancouver and Toronto feature prominently; in addition to London, four other UK cities, including Manchester, feature further down the list, as does both the west coast (Los Angeles and San Francisco) and east coast (New York, Boston) of the US.

Figure 6: Europe and North America: Benchmarking the bohemians.



Figure 7 highlights levels of openness (diversity) across selected European and North American cities. In this case we employ the measure non-western foreign born, particularly as this excludes (for example) Swedes residing in Copenhagen, Americans residing Toronto; which are not the essence of diversity as we attempt to define it. Toronto and Vancouver are significantly more diverse than all the other cities for which we have data – approaching 30 per cent of their populations being born outside Canada and of non-western origin. New York is closely followed by London, which is the most diverse European city. By comparison, the Scandinavian cities tend to display quite low levels of openness as we measure it here.

Turning to an example of comparative bivariate analysis, Figure 8 shows the association between the tech-pole index (relative employment in technology- intensive manufacturing and service sectors), and the relative concentra-tion of bohemians across the European research partners and also for North America. It will be recalled from the earlier discussion of the creative class thesis that a tenet of this is the attraction of technologists and knowl-edge workers to locations that possess a wider creative or artistic milieu – the mechanisms involved in this are developed much more extensively by Wojan, Lambert and McGranahan (2007), and also by Florida, Mellander and Stolarick (2008). This association is shown to be stronger within North

Figure 7: Comparing ‘openness’ across European and North American cities.



America. It is plausible to argue that one reason for this is the greater inten-sive high-technology activity in more cities in the US and Canada, where the greater distances between cities may also mean more representation of both technology entrepreneurs and bohemians in the same places. In Europe, by contrast, bohemians tend to concentrate in the capital city. Conversely, technology entrepreneurship is typically less common therein, but may frequently be found in smaller university cities, for example Oxford and Cambridge in the UK, Lund and Uppsala in Sweden, Oulu and Turku in Finland, Aalborg in Denmark, Twente and Nijmegen in the Netherlands.

Conclusions and implications for further researchThis article provides an overview of the location of creative knowledge workers (the creative class) within selected European countries, the role that quality of place plays within this, and the associated economic outcomes thereof. We do not suggest that all aspects of a concept as nebulous as quality of place can be perfectly captured by these relatively simplistic statistics; what is possible though is to construct some general indicators, underpinned by theory, as demonstrated. This is augmented by comparative data from North American research. The results are quite complex and uneven, but overall there is some support for the idea that the national context in terms of political economy has a role to play; the distribution of the creative class is generally more even within the more coordinated economies, and the public provision index – a

Figure 8: Tech-pole and bohemians: Comparative analysis.



variable introduced specifically for the European context – shows a largely positive association with the location of these workers. As such, we would conclude that Hall and Soskice’s (2001) ‘varieties of capitalism framework’ is a useful model with which to explore further cross-national differences and their impact on observed patterns within the labour market for creative knowl-edge workers. Results for the hypothesized role of diversity in talent attraction are mixed, and require further ‘unpacking’ with regard to the exact mecha-nisms at work; future qualitative work on locational choice is likely to be the most fruitful way forward here (see Raunio, Sotarauta and Andersen 2007 for examples of how this might be undertaken).

An alternative conceptual way in to these issues of shared versus diver-gent preferences is offered by an analysis of the knowledge bases that dif-ferentiate occupations within an aggregate creative class (Asheim and Hansen 2009). We are not able to emulate that here, but we can at least specify separate models for the creative core, creative professionals, and combined creative class, which do show evidence of divergent results. Finally, association between creative class presence and observed eco-nomic outcome is typically high, but causal linkages are unproven. The area of clearest deviation between North America and European econo-mies here is the much stronger relationship in the former between tech-nology-based employment and the presence of the bohemian occupations, which we suggest is likely to arise due to the greater role that smaller pro-vincial cities with relatively high levels of knowledge economy typically play within Europe.

AcknowledgementsResearch was funded by the European Science Foundation via a European Collaborative Research Project in the Social Sciences entitled ‘Technology, Talent and Tolerance in European Cities: A Comparative Analysis’. The UK-based research was funded by Economic & Social Research Council (Grant No. RES-000-23-0467). We also extend our thanks to Meric Gertler of the University of Toronto for his invaluable input into the research. In conjunction with Tara Vinodrai of the University of Waterloo, he is the author of Figures 6–8. Finally we extend our grati-tude to all the European partners for their various efforts in ensuring the project was interesting and enjoyable for all concerned. The usual disclaimer applies.

ReferencesAsheim, B. T. and Hansen, H. K. (2009), ‘Knowledge Bases, Talents and Contexts:

On the Usefulness of the Creative Class Approach in Sweden’, Economic Geography, 85:4, pp. 425–442.

Clifton, N. (2008), ‘The “creative class” in the UK: an initial analysis’, Geografiska, Annaler Series B, Human Geography, 90:1, pp. 63–82.

Cooke, P., De Laurentis, C., Tödtling, F. and Trippl, M. (2007), Regional Knowledge Economies: Markets, Clusters and Innovation, Cheltenham: Edward Elgar.

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

De Laurentis, C. (2006), ‘Digital Knowledge Exploitation: ICT, Memory Institutions and Innovation from Cultural Assets’, Journal of Technology Transfer, 31:1, pp. 77–89.

DeVol, R., Bedroussian, A. and Kim, S. (2007), Best performing cities 2007: Where America’s jobs are created and sustained, Santa Monica, California: Milken Institute.



Florida, R. (2000), Competing in the Age of Talent: Quality of Place and the New Economy, Pittsburgh, Pennsylvania: R.K. Mellon Foundation.

Florida, R. (2002a), The Rise of the Creative Class, New York: Basic Books.

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

Florida, R. (2002c), ‘The economic geography of talent’, Annals of the Association of American Geographers, 92:4, pp. 743–755.

Florida, R. (2007), Who’s your city?, New York: Basic Books.

Florida, R., Mellander, C. and Stolarick, K. (2008), ‘Inside the black box of regional development – human capital, the creative class and tolerance’, Journal of Economic Geography, 8:5, pp. 615–649.

Gertler, M. S., Florida, R., Gates, G. and Vinodrai, T. (2002), Competing on creativ-ity: placing Ontario’s cities in North American context, a report prepared for the Ontario Ministry of Enterprise, Opportunity and Innovation and the Institute for Competitiveness and Prosperity, Toronto, Ontario: Munk Centre for International Studies, University of Toronto.

Hall, P. and Soskice, D. (2001), ‘An introduction to varieties of capitalism’, in P. Hall and D. Soskice (eds), Varieties of Capitalism: The Institutional Foundations of Comparative Advantage, Oxford: Oxford University Press, pp. 1–68.

Hansen, H. K. and Niedomsyl, T. (2009), ‘Migration of the creative class: evidence from Sweden’, Journal of Economic Geography, 9:2, pp. 191–206.

Higgs, P., Cunningham, S. and Bakhshi, H. (2008), Beyond the creative industries: Mapping the creative economy in the United Kingdom, London: NESTA.

Parr, J. B. (2005), ‘Perspectives on the city-region’, Regional Studies, 39:5, pp. 555–566.

Peck, J. (2005), ‘Struggling with the creative class’, International Journal of Urban and Regional Research, 29:4, pp. 740–770.

Raunio, M., Sotarauta, M. and Vaarst Andersen, K. (2007), ‘Attractive city-regions: Varieties of quality of place’, article presented at the American Association of Geographers Annual Meeting, April, San Francisco.

Wojan, T. R., Lambert, D. M. and McGranahan, D. A. (2007), ‘Emoting with their feet: Bohemian attraction to creative milieu’, Journal of Economic Geography, 7:6, pp. 711–736.

Suggested citationClifton, N. and Cooke, P. (2009), ‘Creative knowledge workers and location in

Europe and North America: a comparative review’, Creative Industries Journal 2: 1, pp. 73–89, doi: 10.1386/cij.2.1.73/1

Contributor detailsNick Clifton is a Senior Lecturer in Regional Development at the Creative Leadership & Enterprise Centre, Cardiff School of Management, UWIC. Nick’s main research interests lie in the fields of regional economics, small business and entrepreneur-ship, networks, business strategy, innovation and creativity.

Contact: Creative Leadership & Enterprise Centre, Cardiff School of Management, University of Wales Institute- Cardiff (UWIC), Colchester Avenue, Cardiff, CF23 9XR, UK.Tel: +44 (0)2920 2041 7159E-mail: [email protected]



Phil Cooke is Research Professor in Regional Economic Development, and founding Director (1993) of the Centre for Advanced Studies, Cardiff University. His research interests lie in studies of biotechnology, regional innovation systems, knowledge economies, entrepreneurship, clusters and networks.

Contact: Centre for Advanced Studies in Social Sciences, Cardiff University, 45 Park Place, Cardiff, CF10 3BB, UK.Tel: +44 (0)2920874945E-mail: [email protected]

CIJ_2.1_art_Cooke_073-090 89CIJ_2.1_art_Cooke_073-090 89 12/10/09 2:00:34 PM12/10/09 2:00:34 PM

CIJ_2.1_art_Cooke_073-090 90CIJ_2.1_art_Cooke_073-090 90 12/11/09 2:32:59 PM12/11/09 2:32:59 PM

Creative knowledge workers and location in Europe and North America a comparative review CIJ 2(1)...

Documents

Transcript of Creative knowledge workers and location in Europe and North America a comparative review CIJ 2(1)...