Urban Stud 2012 Denis Jacob 97 114

49(1) 97114, January 2012

0042-0980 Print/1360-063X Online 2011 Urban Studies Journal Limited

DOI: 10.1177/0042098011402235

Jonathan Denis-Jacob is in the Spatial Analysis and Regional Economics Laboratory, Centre Urbanisation Culture et Socit, National Institute of Scientific Research, University of Quebec, 385 rue Sherbrooke Est, Montral, Qubec, Canada H2X 1E3. E-mail: [email protected].

Cultural Industries in Small-sized Canadian Cities: Dream or Reality?Jonathan Denis-Jacob

[Paper first received, July 2010; in final form, December 2010]

Abstract

This paper looks at the residential location of cultural workers in the smallest Canadian cities, with the primary goal of understanding the factors making some more successful than others in attracting them. The study examines employment in 13 cultural industries in 109 small Canadian urban areas using data drawn from the 2006 Canadian census. Six explanatory factors are put forward and entered into a regression model to explain the location of cultural workers in small places: size, location with respect to metropolitan areas, work structure, amenities, elderly populations and public-sector choices. The results suggest that, beyond industry-specific production processes, the location of cultural workers in small cities is also driven by residential and lifestyle preferences.

industries are regarded as driving forces for the regeneration of post-industrial urban fab-rics (Evans, 2001; Florida, 2008; Hutton, 2009; Pratt, 2009), as well as a means to enhance their attractiveness for mobile professionals and capital (Florida, 2002a; Markusen and King, 2003; Scott, 2004; Zukin, 1995). Others see cultural industries as a means for boosting self-confidence and community empower-ment (Evans and Foord, 2006; Huber et al., 1992). Culture has, in short, emerged as a key component in local development strategies.

However, regeneration and economic devel-opment policies based on cultural industries

Introduction

Over the past decade, cultural industries have attracted much attention from urban researchers. An abundant literature draws upon their potential role in urban economic development (Florida, 2002b; Hall, 2000; Landry, 2000; Markusen and King, 2003; Scott, 2004) and in urban regeneration (Evans 2001; Hutton, 2009; Pratt, 2009). Cultural industries are said to be contributing to urban economies in several ways. First, they con-tinue to grow, contributing to employment creation in urban areas while other sectors are experiencing decline (Scott, 2004). Secondly, and perhaps more importantly, cultural

at Charite-Universitaet medizin on March 11, 2015usj.sagepub.comDownloaded from

98 JONAtHAN DENIS-JACOb

have not been successful in all places. This limited success is attributed, on the one hand, to the exaggerated hope placed in culture-led regeneration (Hall, 2000; Scott, 2004), but also to the fact that cultural industries do not necessarily flourish in all places. Indeed, cul-tural industries remain heavily concentrated in a handful of cities at the top of the urban hierarchy (Hall, 1998; Scott, 1999; Sereda, 2007). Research on the location of cultural industries has traditionally focused on the largest metropolitan areas. Yet recent stud-ies have shown that some small cities also exhibit high numbers of creative industries and workers (Hills Strategies, 2006; Nelson, 2005; Petrov, 2007; Power, 2002). However, these studies are either descriptive in nature or look at creative occupations in general, without differentiating between cultural/creative sectors. More research is needed to understand the rationale behind the location of cultural workers employed in different sec-tors in small cities.

This paper looks at workers in 13 cultural industries in small Canadian urban areas, using employment data drawn from the 2006 Canadian census, with the aim of increasing our understanding of their location factors in these places. Six explanatory factors are considered and tested as possible predictors for the strong presence of cultural workers in small cities: size, location, the work culture, public-sector choices, amenities and elderly populations.

Why Size Matters

Cultural industries are usually thought of as metropolitan functions and evidence shows that they are indeed. In North America, the share of cultural employment in total employ-ment is significantly higher in metropolitan areas than elsewhere (Sereda, 2007). In the UK, London accounts for a disproportionate share of cultural employment (Pratt, 1997a, 1997b). Scott (2000) observed that about half

of cultural workers in the US were found in urban areas with populations over a million, with the majority concentrated in the two lar-gest, New York and Los Angeles. City size can therefore determine cities ability to attract and develop cultural industries.

The concentration of cultural employment in large metro areas must be addressed both at the worker and at the firm levels. At the worker level, the necessity of agglomeration and of being near other talented and creative people is central to the discussion around the role of size (Castells, 1996; Hall, 2000). The uncertainty associated with contractual and freelance employment in the cultural sector is a factor. Since a large proportion of cultural workers are hired as freelancers, on a short-term basis (Scott, 2004), they require being in a place where they can keep abreast of current trends and employment oppor-tunities, including through social network-ing activities (Christopherson, 2002). By the same token, many cultural workers, especially artists, work across industries, making thick employment centres more suitable for their professional needs (Markusen and Schrock, 2006). The attraction of particular lifestyles and amenities constitutes another factor explaining the agglomeration of workers in metro areas (Sassen, 1994). At the firm level, explanations lie in the nature of their organi-sational structure. Many cultural industries are characterised by flexible specialisation (Storper and Christopherson, 1987; Shapiro et al., 1992; Scott, 1999), an organisational structure centred around a web of small, independent and highly specialised firms dealing with non-standardised production, constantly interacting with one another and able to adjust rapidly to changes in their industry. Central to this model is the role of outsourcing and contractual production (Storper and Christopherson, 1987). Many traditionally vertically integrated organisa-tions, such as those in broadcasting and pub-lishing, now externalise a considerable part


CULtURAL INDUStRIES IN CANADA 99

of their production to smaller independent firms, a practice that allows increased flexibil-ity, specialised expertise and reduced produc-tion costs. This outsourcing process has led to a substantial growth in the number of small and specialised producers organised around competitivecomplementary relationships (Scott, 2004), resulting in further concentra-tion in metro areas. For many small cultural firms, some degree of spatial proximity is essential because of the weight of specialised labour, tacit knowledge, face-to-face contacts and access to non-codified information. Individual skills in cultural industries are often the results of experience learning and knowledge spill-over processes (Wenting, 2008), both of which usually require personal interactions.

Finally, the fall of transport and communi-cation costs over the past decades has acceler-ated the concentration of cultural production in the largest metro areas (Krugman, 1991; Sassen, 1994). The location choice of many economic activities is the result of a trade-off between economies of scale (from the con-centration of production in one place) and the cost of transporting the output (Polese and Shearmur, 2005). For cultural industries, for which outputs travel at almost no cost across space, the largest metro areas become the optimal locations, allowing them fully to realise economies of scale.

In short, the odds are clearly stacked against small places. However, an emerging literature looks at so-called cultural-creative clusters beyond the metropolis using quantitative (Gibson and Connell, 2004; Hills Strategies, 2006; Nelson, 2005; Petrov, 2007; Power, 2002; Waitt, 2006; Wojan, 2006) and qualitative case study approaches (Evans and Foord, 2006; Gibson and Connell, 2004; Waitt and Gibson, 2009). While most authors generally acknowledge the overwhelming dominance of large metropolitan areas in cultural produc-tion, they all point to the rise of some small cities in cultural/creative industries.

Why Cultural Employment in Small Cities?

The location of industries in small urban areas has traditionally been associated with the so-called crowding-out effect, a progres-sive out-migration (from the metropolis) of economic activities (mostly manufacturing and back-office activities) seeking out more affordable land and labour costs in mid- and small-sized cities (Henderson, 1997). Yet, there has been no evidence to suggest that high costs are systematically pushing cul-tural industries out of major metro areas as they remain the primary places for cultural production. Other factors, beyond urban size and production costs, should be considered to explain the presence of cultural workers beyond the metropolis.

First, size together with location within a certain threshold around large metropolitan areas (100150 km) has proven meaningful in explaining the location patterns of economic activity in Canada and beyond (see Polse and Shearmur, 2004; Polse and Champagne, 1999). For many economic activities, small size and proximity to the metropolis are a double advantage which permits both reduced oper-ating costs (relative to the metropolis) and easy access to metropolitan business functions (Henderson, 1997). For cultural workers how-ever, location with respect to metropolises is not merely about production and transport costs, but more about the very nature of work and organisational structures. A significant number of cultural firms are characterised by flexible organisational structures where individuals build their own schedule and do not necessarily work from a fixed location. Sectors such as the arts and film and video production come to mind as activities which can partly be produced from anywhere given their low transport cost. However, although a cultural worker/firm may produce and create from an isolated location, chances are slim to get new contacts in the same place.



This is where proximity to the large metro area comes in. For those individuals taking advantage of this flexibility and locating in non-metropolitan settings, location on the edge of the metropolis (say within a 100150 km radius) makes sense as frequent face-to-face interactions with clients, partners and institutions remain possible. Centrally located small urban areas are therefore more likely to capture those do-not-want-to-be-in-the-metropolis cultural workers.

Yet location creates both opportunities and disadvantages for small cities, depending on how far away or close enough they are, rela-tive to the large city. Waitt and Gibson (2009) found that Wollongong in Australia has been unsuccessful in developing a vibrant cultural economy due in part to its close proximity to Sydney, which keeps attracting most of Wollongongs cultural workers and consumers. Conversely, for small peripheral cities, distance is a handicap as they often lack the market to sustain year-round cultural activities, being too far from metropolitan areas (Evans and Foord, 2006). For instance, peripheral cities such as Inverness (Scotland) or Timmins (Ontario) suffer from their remote location, being unable to attract metropolitan audiences to local cultural events and activities.

A second possible factor is the absence of a blue-collar legacy. Cultural industries constitute a relatively new type of activities in that they deal mainly with aesthetic and semiotic dimensions (Scott, 2004). Their work and organisational culture is hence different from that of resource-based, construction and heavy manufacturing industries. This suggests that, in cities with a strong blue-collar culture, developing skills and interests that are suitable for the cultural sector may not be an easy task. Waitt and Gibson (2009) argue that Wollongong, Australia, failed to become a vibrant cultural production centre partly because of the weight of industries such as mining and manufacturing in the local economy. The masculine culture prevailing

in Wollongong perceived culture as being soft and associated with leisure and enter-tainment. Similarly, Middleton and Freestone (2008), in a study of culture-led regeneration strategies on local identity in Newcastle, found that the local population, a significant part of whom are blue-collar workers, lacked interest and felt disenfranchised about them.

Thirdly, public-sector choices can be another factor. In Canada and elsewhere, governments and public agencies play a major role in the cultural sector. Several industries, such as the arts, heritage institutions, TV and radio broadcasting as well as motion pic-ture, video and music production are either heavily subsidised by or organised around public institutions. In small capital cities in particular, government expenditures in cul-tural infrastructures and activities can play a major role in the local economy (Coish, 2004; Nelson, 2005; Petrov, 2007). Because public cultural institutions are usually located in capital cities, industries such as the arts and related services and heritage institutions can be expected to employ a higher than average percentage of people in small cities with a capital status. Furthermore, some small cap-ital cities in Canada are the largest city in their province as well as the administrative centre (Charlottetown, Prince Edward Island, is an example). These urban areas benefit from their central place role in cultural production, making them the only place where cultural production genuinely takes place in the province. Also, the occupational structure of capital cities is generally different from that of non-capital cities (Carroll and Meyer, 1982). Capital cities usually have a higher share of professionals and service-sector workers with higher wages and a taste for culture. By the same token, location choices of public cor-porations such as the Canadian Broadcasting Corporation (CBC-SRC) can benefit small cities. Several small-sized urban areas are the home of a CBC/SRC radio or television chan-nel to serve local markets.



Fourthly, the presence of certain types of urban and natural amenities may also influence the location of cultural workers. Considerable attention has been paid to amenity-based location choices (Gibson, 2002; Florida, 2002b; Heilburn, 1996; Lewis and Donald, 2010; Markusen and Schrock, 2006; Nelson, 2005). Although much of the discourse on amenities in recent years is related to the creative class theory and has focused primarily on large metropolitan areas, amenities are increasingly a factor explaining location choices in small cities. Amenity and lifestyle factors have even proven meaningful in explaining the rapid growth of some small Ontario towns around Toronto (Wilkinson and Murray, 1991; Dahms and McComb, 1999). It is argued that cultural and creative workers prefer (and can afford) living in attractive natural environments (i.e. coastline and lakefront, mountain landscapes) and in those with a small-town atmosphere (traditional urban fabrics and vibrant down-towns). From this perspective, location in small urban areas would be based on lifestyle preferences rather than merely on industry imperatives (Dahms, 1998; Gottlieb, 1994), choices made possible by increased mobil-ity, flexible work practices and electronic communications (Dahms, 1998; Markusen and Schrock, 2006; van Oort et al., 2003). Although not all cultural workers benefit from flexible conditions, more do than in other industries. Many designers, artists and writ-ers, especially those who are self-employed, can easily produce from anywhere. Amenities, however, can hardly be detached from prox-imity to large metro areas. A place may be beautiful but will not become a desirable residential location if remote because, as dis-cussed earlier, proximity to large metro areas still matters. Yet not all small towns located around metro areas are cultural hotspots. This is where amenities come in. Centrally located places with amenities may therefore prove attractive locations for a number of cultural

workers. This recalls Friedmann (1973) and his urban field concept, an ecological unit comprised within a 150160 km radius from metro areas where residential settlement takes place based on lifestyle, employment and mobility. Within the urban field, people seek out locations where they can both cre-ate customised residential environments and interact with the metropolis (Dahms, 1998). As well as attraction factors based on amen-ities, the lower cost of living (relative to large metro areas) makes small places appealing to cultural and creative workers (Dahms, 1998; Markusen and Schrock, 2006).

Lastly, a fifth possible factor is the presence of elderly populations which has been put forward as a predictor for cultural consump-tion (Beyers, 2002; Ewoudou, 2005) and growth in small towns and rural areas (Frey, 1993; Dahms, 1998). Elderly, and even more so retirees, have abundant leisure time and, often, financial resources and therefore a greater propensity to consume given cultural activities. Their propensity to produce given cultural products is also greater as activities such as the arts can be produced for both lei-sure and professional purposes. More flexible (and appealing) work practices in cultural industries have led many to extend their professional activities beyond retirement. Furthermore, elderly populations are more likely to locate in non-metropolitan areas being seldom constrained by a job location (Frey, 1993). In addition, elderly populations often play a central role in small town com-munity life, including in heritage preservation and the local art scene. However, there could be an overlap with amenities as location choices after retirement can also be based on them. For instance, Elliott Lake (Ontario) has been successful in attracting important elderly populations in recent years because of its natural attributes. Similarly, Port Hope, Cobourg (Ontario) and Parksville (B.C.) have also become the home of an important retired community because of their amenities.



Study Area, Data and Methodology

Study Area and Data

The study examines all Canadian urban areas (144) with populations above 10 000, classi-fied as census metropolitan areas (CMAs) or census agglomerations (CAs) focusing on small cities (109) with populations below 100 000. The employment data, drawn from the 2006 census, are by the place of residence (not place of work) and therefore account for the number of cultural workers living in small cities. The intended goal was to look at the location patterns of cultural industries in small cities; however, given the limitations of the data available (by place of residence, not place of work), this has not been pos-sible and the residential location of cultural workers was finally retained. The data are organised around 13 cultural industries coded at the 4-digit 1997 NAICS (North American Industry Classification) level. Industries were chosen over occupations in line with a growing literature which refers to cultural production as an industrial production sys-tem (Pratt, 1997a; Power, 2002; Evans, 2001). In effect, cultural production does not merely rely on artists and creative people, but also on many others who are also central to the pro-duction process (managers, technical staff and so on). Occupations, on the other hand, are useful to study work tasks and where specific professional groups tend to locate, but fail to capture fully the total employment in given industries. Therefore, industries were chosen over occupations because they permit one to draw a more accurate picture of the scope of cultural employment in small cities.

Adopting a meaningful definition is a chal-lenge given the definitional debates surround-ing cultural industries (Hesmondhalgh and Pratt, 2005; OConnor, 1999; Pratt, 1997a, 1997b; Scott, 2004). Definitions differ between researchers, depending on their specific aims, and between nations using different national

industry classification systems, making con-sistent international definitions difficult (Bryan et al., 2000). Only cultural industries concerned with the transmission of signs and symbols (Bourdieu, 1971; Hesmondhalgh and Pratt, 2005), those which provide goods and services whose subjective meaning is high in comparison with their utilitarian purpose and those for which the aesthetic content and sign-value to the consumer are important (Scott, 2004) are examined. In other words, the study focuses on industries where the creation of cultural content is central to the value chain. Thirteen cultural industries are selected and grouped into nine sub-sectors to simplify the analysis (Table 1).

The classification is largely inspired by that of Coishs (2004) study of Canadas metropolitan culture clusters whose 17 cultural classes are based on the definition of cultural goods and services proposed by the Canadian Framework for Cultural Statistics (Statistics Canada, 2004). Four classes from Coishs (2004) definition (printing and related support activities,

Table 1. Employment in the cultural industries in Canadian cities, 2006

Cultural industriesEmployment in 144 cities

Book, periodical and music stores

21 010

Newspaper, periodical, book and database publishers

135 745

Motion picture, video and sound recording industries

57 795

Radio and television broadcasting

38 390

Pay TV, specialty TV and programme distribution

25 975

Specialised design services 46 345Advertising and related services 63 800The arts and related services 33 365Heritage institutions 18 925

Total (all cultural industries) 441 350



manufacturing and reproducing magnetic and optical media, information services as well as architectural and landscape architec-tural services) were excluded because they mainly include non-cultural employment. Other business-to-business activities, such as advertising and design, have been included because they are mainly concerned with the creation of value through symbolism and/or aesthetics and because an important part of their value chain includes creative and cultural inputs.

The data are by place of residence, not place of production, therefore data must be interpreted as the location of cultural work-ers. Because some cultural workers may live in a small city but work elsewhere, the data constitute a risk if interpreted as if they related to the place of production. This is not so much of an issue for large metropolitan areas or small peripheral urban areas, but can be for small cities located near urban centres. For that reason, the study looks at the residential location of workers employed in cultural industries as opposed to the distribution of cultural production.

The data feature some limitations. First, industrial classification systems are not necessarily well suited to the identification of cultural employment (Evans, 2001; Scott, 2000). Many features of cultural industries (part-time, contractual and free-lance employment, multiple job occupancy, multiple job locations, home-based employ-ment and so on) are difficult to capture adequately via industry classification sys-tems. Furthermore, many cultural sectors overlap with non-cultural activities, notably advertising (with public relations activities),1 publishing (with database publishing) and heritage institutions (with zoological and botanical gardens and amusement parks). Zoological and botanical gardens and amuse-ment parks are arguably leisure activities and more scientific than some of cultural produc-tion already broadly defined.

Methodology

Both descriptive statistics (locations quotients) and regression models are employed to assess the role of each factor on the strong presence of cultural workers in small cities.2 Following the presentation of the location quotients (per industry and city-size class) these are then entered as dependent variables in the regressions. Nine regression models were built (one for all cultural industries and one for each cultural industry) and applied to the 109 urban areas with populations under 100 000.

Descriptive Analysis: Small Cities with Big Cultural Numbers

Figure 1 both confirms and questions the role of city size for cultural industries. The statis-tical relationship between cultural employ-ment (location quotient: all industries) is positive, but the R2 is fairly low, leaving ample room for other explanations.3 We can see that some of the smallest cities exhibit a similar (or even higher) location quotient than the largest metro areas (Toronto, Montreal and Vancouver). Canmore (Alberta) and Stratford (Ontario) are the most notable cases. With a population of about 30 000, both cities exhibit the highest location quotients in total cultural employment with 1.65 and 1.51 respectively. Other cities under 30 000 also have relatively high scores for their size; Port Hope, Ontario (1.06),Whitehorse, YT (1.00), Nanaimo, BC (0.97), Elliot Lake, Ontario (0.97) and Yellowknife, NWT (0.95). In comparison, the large metro areas of Quebec City and Calgary display scores of 0.94 and 0.92 respectively. The limited R2 (0.201) tells us that we should look to explanations beyond city size, to which we now turn.

In Tables 2 and 3, we employ a grouping technique taken from Polese and Shearmur (2004) where urban areas are grouped based on city size and distance to the closest Top 8 census metropolitan areas (CMA).4 Four groups of cities are created and a location



quotient is calculated for each one. Distance is calculated with a distance matrix on the road network using GIS. Cities are central when within 200 km of, and peripheral when beyond 200 km from, a Top 8 metro area. Small cities are defined as those with a popu-lation below 100 000 and the smaller cities as those under 30 000 people.

The 200 km cut-off and the population thresholds have been determined for three main reasons. First, a minimum number of 20 urban areas and a minimum population of 400 000 per group were needed to ensure the accuracy of the analysis. Too few observations in each group would have given too much weight to specific cities with extreme scores. In addition, the 30 000 threshold permitted

a relatively even distribution between groups of above 30 000 (50 units) and below 30 000 (59 units). Secondly, the 200 km cut-off takes into account the reality of occasional commuting and travel patterns. On the road system, depending on driving conditions, 200 km correspond to a two-hour journey to or from the metropolis. This distance permits commuting to the metropolis on an irregular basis. Thirdly, tests have been made for 150 km, 200 km, 250 km and 300 km as well as for different sizes. The population threshold of 30 000 and the distance cut-off of 200 km have provided the most meaningful results.

Table 3 shows that small urban areas have higher than average location quotients in only three sectors (heritage institutions, the

Figure 1. Relationship between cultural specialisation and size. Notes: Pearson correlation coefficient: 0.461 (significant at the 0.01 level); regression analysis: R = 0.201; adjusted R2 = 0.195.



arts and related services and radio and TV broadcasting). Heritage institutions and the arts and related services exhibit particularly high LQs in central cities with a popula-tion below 30 000 people. Proximity to the metropolis provides cultural workers (and organisations) with easy access to a diversity of resources they may not find in a small city (Canmore, Alberta, and Port Hope, Ontario, are examples). On the other hand, the LQ for employment in radio and TV broadcasting is higher in peripheral cities with a population below 30 000. This industry, unlike PAY TV, depends on local content, including local news and advertisements, and therefore requires proximity to local communities. In contrast, employment in PAY TV is concen-trated in Toronto and Montrealand almost non-existent in small urban areasas it relies on national subscriptions. Peripheral cities such as Whitehorse (YT), Yellowknife (NWT) and Rimouski (Quebec) clearly benefit from the protection effect of distance with location quotients of 2.3, 2.5 and 1.6 respectively.

The Regression Models

Nine regression models were built using SPSS where the dependent variable is a loca-tion quotient per industry. The models are performed for all 109 census agglomerations (CAs) with populations below 100 000 but over 10 000. The explanatory factors already discussed are expressed via six independent variables (including logged city size). The operationalisation of these factors is the main challenge and is discussed further.

The median housing value in 2005 is used as a proxy for the attractiveness of a citys urban and natural environment, therefore for the presence of amenities. The presence of natural amenities such as mountains, lakes and forests or urban amenities such as a well-preserved historical town centre and a small-town character increasingly constitutes a powerful predictor for high land values (Clark, 2000).We implicitly assume that (urban and natural) amenities are capitalised in housing values. In Canada, unlike in the US, the quality of

Table 3. Location quotients of cultural industries, by synthetic groups of small urban areas

Cultural SectorsSmall central

Small peripheral

Very small central

Very small peripheral

All cultural industries 0.56 0.55 0.70 0.54Books, periodical and music stores 0.88 0.72 0.69 0.94Publishing 0.60 0.73 0.79 0.59Motion picture, video and sound 0.36 0.38 0.38 0.41Radio/TV broadcasting 0.42 0.75 0.52 1.16Pay TV 0.68 0.51 0.53 0.47Advertising and design 0.50 0.37 0.44 0.29The arts and related services 0.56 0.40 1.07 0.40Heritage institutions 1.01 0.80 1.63 0.97

Table 2. Synthetic groups of small urban areas based on city size and distance

Synthetic regions: urban areas

Population threshold

Distance from Top 8 metro (km)

Total population

Employment Spatial units

Small central 100 00031 000 Within 200 1 325 908 646 190 23Small peripheral 100 00031 000 Beyond 200 1 505 201 758 940 27Very small central Below 31 000 Within 200 418 731 205 215 23Very small peripheral Below 31 000 Beyond 200 658 625 320 935 36



public services (schools, hospitals and police forces) plays almost no role in property values as they rarely vary in quality from one place to another, being run and/or funded by prov-incial governments. Variations in property values in small cities are most likely to be related to the presence of specific amenities highly sought out by residential populations.

A correlation analysis has been performed and suggests that size and median housing values are not correlated for urban areas below 100 000 (see Table 4). Except for a few cases (such as Woods Buffalo, Alberta, and Yellowknife, NWT) where property prices are higher because of characteristics of the local economy (the presence of natural resources and government functions), most other urban areas with high median housing values are indeed known for their residential attractiveness. Note that distance from the Top 8 metro areas and housing values are negatively correlated, confirming that the residential attractiveness of small cities is tied to proximity to large metro areas. The logged distance in km from the nearest Top 8 metro areas, calculated on the road network using GIS, is used to determine the role of location with respect to major metro areas.

The percentage of blue-collar workers is used to assess the effect of the so-called class legacy on cultural employment. This variable captures blue-collar occupations from the National Occupational Classification (NOC-S) 2006. Occupations have been chosen over

industries to isolate blue-collar workers from white-collar occupations in the same industries. Some cities may have similar-sized employment numbers in one industry, but with different occupational structures (blue collars vs managers in the pulp mill industry, for instance). Occupations are therefore better suited for assessing the impact of the so-called blue-collar work culture. Small urban areas with lower shares of employment in blue-collar occupations are expected to score higher than those with high blue-collar employment numbers in terms of cultural employment. Peripheral urban areas would be expected to rely more on blue-collar occupations than those near large metropolitan areas given their depend-ence on natural resources. However, distance and the percentage of blue-collar workers are not strongly correlated (Table 5).

The percentage of population aged 65 years and older is used as a proxy for the presence of an important retired population. Although we could expect some circularity between elderly populations and amenities, the correlation analysis confirms a weak relationship, sug-gesting that not all amenity-rich towns are retirement communities.

A dummy is used as a control variable for the four capital cities with populations below 100 000 residents (Fredericton, Charlottetown, Whitehorse, Yellowknife). A dummy variable is used for the presence of the CBC/SRC station in the model for radio and TV broadcasting because it plays a central role in this industry.

Table 4. Correlation analysis on independent variables (N = 109)

Correlation coefficients 1 2 3 4 5 6 7

1 Total population (2006) 1.00 0.15 -0.04 -0.07 0.09 -0.01 -0.032 Median housing value ($) 0.15 1.00 -0.17 -0.28** 0.07 -0.13 -0.023 CBC/SRC dummy -0.04 -0.17 1.00 0.37** 0.32** -0.29** -0.29**4 Distance (km) -0.07 -0.28** 0.37** 1.00 0.17 -0.26** -0.145 Capital cities 0.09 0.07 0.32** 0.17 1.00 -0.25** -0.36**6 Elderly (percentage) -0.01 -0.13 -0.29** -0.26** -0.25** 1.00 -0.067 Blue-collar workers (percentage) -0.03 -0.02 -0.29** -0.14 -0.36** -0.06 1.00

Note: ** significant at 0.01.



Results

The results suggest relatively different R values between industries. The most robust models (as measured by the adjusted R2) are those of the arts and related services (0.421), all cultural industries (0.360), radio and TV broadcasting (0.274) and advertising and design (0.210). Results show that the six variables have little effect on the location of workers in motion picture, video and sound recording and pay TV as they are mainly concentrated in major metro areas and almost non-existent in small cities. Surprisingly, the models for heritage institu-tions and book, periodical and music stores are not robust, despite high employment numbers in small-sized urban areas. Let us now turn to the regression coefficients for each variable (Table 6).

Size is only significant for advertising and design and does not appear to play a role in any other sector. Distance with respect to large metro areas is not significant in any model. As discussed earlier, location with respect to metropolitan areas may have contradictory

effects depending on the attributes of cultural goods or services. We observed that urban areas with similar locations (central or per-ipheral) exhibited quite dissimilar cultural specialisation scores. Examples are Stratford and Ingersoll in Ontario. Both towns lie approximately 150 km south-west of Toronto. The former is the second most specialised in cultural employment, whereas the latter has the lowest score of any urban areas. The same holds true for peripheral cities. Whitehorse and Yellowknife exhibit high scores in several sectors, while Thompson (Manitoba) is at the very bottom in all rankings. Location matters for attracting cultural workers, but is seldom sufficient to ensure success.

The role of elderly populations is significant for: all cultural industries, publishing, motion picture, video/sound recording and the arts and related services. The findings for these industries suggest that elderly populations have an effect on industries characterised by flexible specialisation, freelance employ-ment and those which can be produced for both leisure and professional purposes. Retirement centres such as Elliot Lake, Port

Table 5. Correlation between independent and dependent variables (correlations are presented in the same order as in table 4)

Dependent variable (LQ)

Independent variables

1 2 3 4 5 6 7

Total cultural employment 0.007 0.430** 0.120 -0.148 0.274** 0.197* -0.388** Book, periodical and music stores

0.034 0.094 0.110 0.050 0.217* 0.176 -0.251*

Newspaper, periodical, book and database publishing

0.031 0.177 -0.004 -0.091 0.074 0.323** -0.195*

Motion picture, video, sound recording

-0.079 0.101 0.104 0.040 0.206* 0.035 -0.180

Radio and TV broadcasting -0.149 -0.031 0.493** 0.301** 0.260** -0.097 -0.325** Pay TV, specialty TV and programme distribution

0.118 0.131 0.059 -0.016 0.006 0.133 -0.230*

Design and advertising 0.208* 0.388** -0.118 -0.280** 0.057 0.161 -0.109The arts and related services 0.010 0.315** -0.020 -0.237* 0.178 0.132 -0.177Heritage institutions -0.091 0.363** -0.023 -0.143 0.096 -0.150 -0.109

Notes: *significant at 0.05; **significant at 0.01.



Hope, Cobourg and Collingwood (Ontario) all exhibit high scores of cultural employ-ment. This confirms the hypothesis that elderly populations have a greater ability to engage in cultural industries in small cities because of their greater liberty to live outside metro areas.

The results suggest that small capital cit-ies are more likely to have a higher share of cultural employment, consistent with expectations. Being a capital is a predictor for all cultural industries, motion picture, video/sound recording, the arts and related services and heritage institutions. The pres-ence of public cultural institutions (concert halls, museums, art galleries) in the small capital cities of Whitehorse, Yellowknife, Charlottetown and Fredericton pushes up cultural employment. Yet public expendi-tures in the arts and culture are not the only reasons why these cities specialise to a

greater extent in cultural industries. These four cities are also the central place of their province. These urban areas exhibit high spe-cialisation scores in the visitor-dependent sectors (the arts and heritage institutions) because they are often the only location where they take place in their province (Whitehorse, Yellowknife and Charlottetown are examples).

The Canadian Broadcasting Corporation/ Socit Radio-Canada (CBC/SRC) dummy is the only significant one in the model for radio and TV broadcasting, confirming its central role in the specialisation of small peripheral cities in this industry. It should be empha-sised, however, that most urban areas with a CBC/SRC station are central places in their respective region. TV and radio broadcasting is a sector which requires a certain degree of proximity with local communities because of local news and advertisements. Although

Table 6. Regression models: summary

Variables: Standardised coefficients

Industry N Adjusted R2

Log distance

Log size Housing value

Capital dummy

Percentage of blue collars

Percentage of elderly

CBC dummy

All cultural industries

107 0.360 0.012 -0.066 0.349** 0.261** -0.287 0.334** n.a

Book, periodical and music stores

105 0.045 0.134 0.068 0.177 -0.007 -0.193 0.146 n.a

Publishing 107 0.131 0.011 -0.046 0.241** 0.140 -0.129 0.319** n.aMotion picture, video and sound recording

107 0.092 0.121 -0.114 0.105 0.240* -0.089 0.271** n.a

Radio and TV broadcasting

109 0.274 0.124 -0.132 0.096 n.a -0.187 0.062 0.419**

Pay TV 106 0.006 0.050 0.026 0.093 -0.093 -0.255 0.025 n.aAdvertising and design

108 0.210 -0.144 0.201* 0.329** 0.091 -0.060 0.154 n.a

The arts and related services

105 0.421 -0.008 0.016 0.349** 0.428** -0.120 0.424** n.a

Heritage institutions

108 0.028 -0.035 -0.025 -0.022 0.315** 0.179 0.017 n.a

Notes: Outliers have been removed from some models because of their extreme values. ** significant at 0.01; * significant at 0.05.



production aimed at national audiences is centralised in Toronto and Montreal for the most part (for example, national news and entertainment programmes), location in peripheral urban areas is essential to ensure local content for these communities. However, not all peripheral central places have a CBC/SRC station. Peripheral places such as Vernon, BC, and Estevan, Saskatchewan, are central places in their region but have no or little employment in this industry. Other cities have benefited from the location choices of the CBC/SRC.

The class legacy explanation cannot be rejected for cultural industries. The results suggest that cities with high shares of employ-ment in blue-collar occupations are less spe-cialised in all cultural industries, radio and TV broadcasting and pay TV. Cultural work-ers in general are hence less likely to locate in places with a strong blue-collar work culture. Canmore (Alberta), Elliot Lake (Ontario) and Owen Sound (Ontario) are examples of places with lower than average blue-collar occupa-tion numbers and high cultural employment figures. In contrast, urban areas with very high numbers in blue-collar occupations such as Woods Buffalo (Alberta) and Estevan (Saskatchewan) have few cultural workers. Surprisingly, the variable is not significant for other core creative sectors such as the arts and related services and motion picture, video/sound recording. Housing value (a proxy for amenities) is significant for all cul-tural industries, publishing, advertising and design, and the arts and related services. This confirms the footlooseness of cultural work-ers in given sectors and their preference for amenity-rich environments, probably because of their flexible work conditions and ability to telework. This also suggests that workers in these sectors can afford higher land values in amenity-rich communities.

Figure 2 shows that most small urban areas with location quotients near or above 1 also have high housing values. Cities such

as Stratford, Port Hope, Centre Wellington, Collingwood, Cobourg and Tilsonburg are known to be attractive places because of their small-town atmosphere and well-preserved urban fabric. Similarly, places like Canmore, Nanaimo, Parksville and Owen Sound are considered highly desirable places in which to live because of their natural amenities. For example, location in Collingwood, Centre Wellington, Parksville and Port Hope has proven suitable for design and advertising workers (and firms) as they offer both a pleasant place to live, work and play, and proximity to corporate headquarters in Vancouver and Toronto.

Workers in the arts and related services follow a similar logic. Many small towns, including Stratford, Cobourg (Ontario) and Canmore (Alberta), are effectively the homes of Canadian artists and cultural personalities. Stratford, Ontario, is a notable example of an amenity-rich town which continues to attract cultural workers. The town, located half-way between Toronto and Detroit, is famous for its well-preserved historical town centre and its Shakespeare Festival. With a location quo-tient of 6.1 in workers in the arts and related services, the town is the most specialised of any Canadian city. It also has above average scores in book, periodical and music stores and publishing. Canmore (Alberta) is another interesting case. The town lies an hour and a half from Calgary and is the gateway to the Banff National park. Its beautiful natural setting has made it one of the most desirable residential locations in the country, especially for those passionate about outdoor activi-ties (skiing and hiking in particular). The town scores high in all cultural industries, book, periodical and music stores, publish-ing, the arts and heritage institutions. Wood Buffalo (Alberta), Whitehorse (Yukon) and Yellowknife (Northwest Territories), on the other hand, are not amenity-based residen-tial locations but places where property values are driven by the local economic base



(dependent on government services and/or natural resources).

Conclusion and Discussion

This paper examined the residential location of cultural workers in small Canadian urban areas. The results suggest that the strong pres-ence of cultural workers is clearly not merely a matter of size. While most cultural workers remain concentrated in major metropolitan areas, some small cities are also successful in attracting them. Small places such as Stratford (Ontario), Canmore (Alberta), Port Hope (Ontario) and Nanaimo (BC) have indeed a high share of their working population employed in cultural industries. The six

explanatory factors exhibit great variability depending on the industry. However, for all cultural industries taken as a whole, the presence of amenities, the absence of a blue-collar work culture and the presence of a large elderly population are positive predictors of large cultural worker populations in small urban areas. Being a capital city is also an advantage as capital cities benefit from the presence of government cultural institutions.

The results also suggest that location in non-metropolitan places is facilitated by the flexible organisational and work structures of some cultural industries. For workers employed in industries such as publishing, advertising and design and the arts and related activities, location decisions appear to be

Figure 2. Relationship between median housing value and cultural specialisation in small cities. Notes: Pearson correlation coefficient: 0.496 (significant at 0.01).



made based on residential preferences rather than merely on production imperatives. Hence, the presence of (urban and natural) amenities appears to be a selling point for many cultural workers locating in small cities. Moreover, elderly populations play a central role in the cultural sector as a whole, as well as in publishing, motion picture, video/sound recording and the arts and related services by being willing and able to locate outside metro areas. Although not significant, proximity to the metropolis remains essential to maintain a link with a workplace, clients or/and part-ners, as most small cities with high scores are located within 200 km of a Top 8 metro area.

These findings are in line with current discussions around the rise of a residential economy (Davezies, 2009), where location decisions are increasingly made based on life-style and residential preferences made possible by an increasingly footloose population. In effect, the results raise several questions about the increasingly unclear distinction between places of work and places of residence in the cultural economy. With such flexible and unstable work conditions in many indus-tries, the very notion of localised production becomes fuzzy. While it is still obvious that large metropolitan areas remain the main nodes of cultural production, the rise of small cities as sites of cultural workers residence and, potentially, cultural production, confirms the emergence of new forms of relationship between work, production, leisure and living.

This, however, remains difficult to capture with this study and therefore further quantita-tive and qualitative research would be needed. The use of quantitative data by place of pro-duction would determine the extent to which cultural production genuinely takes place outside metro areas. In addition, qualitative research, through interviews with cultural and creative workers living in small cities, could confirm the relevance of the location factors from a personal perspective. Special attention could also be devoted to the nature of their

professional practice (employment status, work schedule) in order to determine whether patterns can be identified, including between industries. Finally, the type and frequency of interaction with the metropolis (number of monthly visits, clients and collaborators, metropolitan resources sought out, etc.) could be further investigated in order to understand the genuine role of small cities in a growing cultural landscape.

Notes

1. Specialised design services and advertising and related services are also put together in all analyses because they arguably constitute high-order services, aiming at firms and companies, rather than the general public. Although their activities are different, we argue that their nature is relatively similar in that they require frequent contacts with and feedback from their clients, deal mostly with custom-made production and are concerned with the creation of value through symbolism and/or aesthetics.

2. The location quotient (LQ) is a measure of specialisation of a citys share of employment in a given industry relative to the national average. When above 1, specialisation is higher than the national norm, at 1 it is equal and below 1 it is lower.

3. A regression analysis has been used to test the relationship between size and cultural specialisation. The natural logarithm has been used for both city size (total population in 2006) and cultural specialisation (LQ for all cultural industries) to correct for the imbalance in variable distributions.

4. The top eight metropolitan areas are Toronto, Montreal, Vancouver, OttawaGatineau, Calgary, Edmonton, Quebec City and Winnipeg.

References

Beyers, W. B. (2002) Culture, services and regional development, The Service Industries Journal, 22, pp. 434.

Bourdieu, P. (1971) Le march des biens symbol-iques, Lanne sociologique, 22, pp. 49126.



Bryan, J., Hill, S., Munday, M. and Roberts, A. (2000) Assessing the role of the arts and cul-tural industries in a local economy, Environ-ment and Planning A, 32, pp. 13911408.

Carroll, G. and Meyer, J. (1982) Capital cities in the American urban system: the impact of state expansion, The American Journal of Soci-ology, 88, pp. 565578.

Castells, M. (1996) The Network Society. London: Blackwell.

Christopherson, S. (2002) Project work in con-text: regulatory change and the new geography of media, Environment and Planning A, 34, pp. 20032015.

Clark, T. N. (2000) Old and new paradigms for urban research: globalization and the fiscal austerity and urban innovation project, Urban Affairs Review, 36(1), pp. 345.

Clark, T. N., Lloyd, R., Wong, K. K. and Jain, P. (2002) Amenities drive urban growth, Journal of Urban Affairs, 24(5), pp. 493515.

Coish, D. (2004) Census metropolitan areas as culture clusters: trends and conditions in census metropolitan areas. Analytical paper, Culture, Tourism and the Centre for Education Statis-tics, Statistics Canada, Ottawa, ON.

Dahms, F. (1995) Dying villages, counterurbani-zation and the urban field: a Canadian per-spective, Journal of Rural Studies, 11, pp. 2133.

Dahms, F. (1998) Settlement evolution in the are-na society in the urban field, Journal of Rural Studies, 14, pp. 299320.

Dahms, F. and McComb, J. (1999) Counterur-banization, interaction and functional change in a rural amenity area: a Canadian example, Journal of Rural Studies, 15, pp. 129146.

Davezies, L. (2009) Lconomie locale rsidentielle, Gographie, conomie, socit, 11(1), pp. 4753.

Evans, G. (2001) Cultural Planning: An Urban Renaissance? New York: Routledge.

Evans, G. and Foord, J. (2006) Small cities for a small country, in: D. Bell and M. Jayne (Eds) Small Cities: Urban Experience beyond the Metropolis, pp. 151167. New York: Routledge.

Ewoudou, J. (2005) Understanding culture con-sumption in Canada. Research paper, Culture, Tourism and the Centre for Education Statis-tics, Statistics Canada, Ottawa, ON.

Florida, R. (2002a) Bohemia and economic geogra-phy, Journal of Economic Geography, 2, pp. 5571.

Florida, R. (2002b) The Rise of the Creative Class: And How its Transforming Work, Leisure, and Everyday Life. New York: Basic Books.

Florida, R. (2008) Whos Your City? How the Crea-tive Economy is Making Where to Live the Most Important Decision of Your Life. New York: Basic Books.

Frey, W. H. (1993) The new urban revival in the United States, Urban Studies, 30, pp. 741744.

Friedmann, J. (1973) The urban field as human habitat, in: L. S. Bourne and J. W. Simmons (Eds) Systems of Cities, pp. 4252. Toronto: Oxford University Press.

Gibson, C. (2002) Rural transformation and cul-tural industries: popular music on the New South Wales far north coast, Australian Geo-graphical Studies, 40, pp. 337356.

Gibson, C. and Connell, J. (2004) Cultural indus-try production in remote places: indigenous popular music in Australia, in: D. Power and A. Scott (Eds) Cultural Industries and the Pro-duction of Culture, pp. 243258. New York: Routledge.

Gottlieb, P. (1994) Amenities as an economic development tool: is there enough evidence?, Economic Development Quarterly, 8, pp. 270285.

Hall, P. (1998) Cities in Civilization. New York: Pantheon.

Hall, P. (2000) Creative cities and economic development, Urban Studies, 37, pp. 639649.

Heilburn, J. (1996) Growth, accessibility and the distribution of arts activity in the United States: 1980 to 1990, Journal of Cultural Eco-nomics, 20, pp. 283296.

Henderson, J. V. (1997) Medium sized cities, Regional Science and Urban Economics, 27, pp. 583612.

Hesmondhalgh, D. and Pratt, A. (2005) Cultural industries and cultural policy, International Journal of Cultural Policy, 11, pp. 113.

Hills Strategies Research Inc. (2006) Artists in small and rural municipalities in Canada, Statistical Insights on the Arts, 4(3) (http://www.hillstrategies.com/resources_details.php?resUID=1000336&lang=0).

Huber, M., Williams, A. and Shaw, G. (1992) Cul-ture and economic policy: a survey of the role of local authorities. Working Paper No. 5, Tour-ism Research Group, Department of Geogra-phy, University of Exeter.



Hutton, T. (2008) The New Economy of the Inner City: Restructuring, Regeneration and Dislo-cation in the Twenty-first-century Metropolis. New York: Routledge.

Hutton, T. (2009) Trajectories of the new econo-my: regeneration and dislocation in the inner city, Urban Studies, 46, pp. 9871001.

Krugman, P. (1991) Geography and Trade. Cam-bridge, MA: MIT Press.

Landry, C. (2000) The Creative City: A Toolkit for Urban Innovators. London: Earthscan.

Lewis, N. M. and Donald, B. (2010) A new rubric for creative city potential in Canadas smaller cities, Urban Studies, 47, pp. 2954.

Markusen, A. and King, D. (2003) The artistic div-idend: the arts hidden contributions to regional development. Project on Regional and Indus-trial Economics, Humphrey Institute of Public Affairs, University of Minnesota.

Markusen, A. and Schrock, G. (2006) The artis-tic dividend: urban artistic specialisation and economic development implications, Urban Studies, 43, pp. 16611686.

Middleton, C. and Freestone, P. (2008) The impact of culture-led regeneration on regional identity in north east England. Paper presented at the Regional Studies Association Internation-al Conference: The Dilemnas of Integration and Competition, Prague, May.

Nelson, R. (2005) A cultural hinterland? Search-ing for the creative class in the small Canadian city, in: W. F. Garrett-Petts (Ed.) The Small Cit-ies Book: On the Cultural Future of Small Cities, pp. 85109. Vancouver: New Stars Books.

OConnor, J. (1999) The definition of cultural industries. Manchester Institute for Popular Culture, Manchester.

Oort, F. van, Weterings, A. and Verlinde, H. (2003) Residential amenities of knowledge workers and the location of ICT-firms in the Netherlands, Tijdschrift voor Economische en Sociale Geografie, 94(4), pp. 516523.

Petrov, A. (2007) A look beyond metropolis: exploring creative class in the Canadian pe-riphery, Canadian Journal of Regional Science, 30(3), pp. 451474.

Polse, M. and Champagne, E. (1999) Location matters: comparing the distribution of eco-nomic activity in the Mexican and Canadian urban systems, International Regional Science Review, 22(1), pp. 102132.

Polse, M. and Shearmur, R. (2004) Is distance really dead? Comparing industrial location patterns over time in Canada, International Regional Science Review, 27, pp. 431457.

Polse, M. and Shearmur, R. (2005) conomie ur-baine et rgionale: introduction la Gographie conomique, 2nd edn. Paris: Economica.

Polse, M., Rubiera-Morolln, F. and Shearmur, R. (2005) Observing regularities in location pat-terns: an analysis of the spatial distribution of economic activity in Spain. Working paper, INRS Urbanisation Culture et socit (http://www1.ucs.inrs.ca/pdf/inedit2005_08.pdf).

Power, D. (2002) Cultural industries in Sweden: an assessment of their place in the Swedish economy, Economic Geography, 78, pp. 103127.

Pratt, A. (1997a) The cultural industries pro-duction system: a case study of employment change in Britain, 198491, Environment and Planning A, 29, pp. 19531974.

Pratt, A. (1997b) The cultural industries sector: its definition and character from secondary sources on employment and trade, Britain 198491. Research Papers in Environmental and Spatial Analysis No. 41, Department of Geography, London School of Economics.

Pratt, A. (2009) Urban regeneration: from the arts feel good factor to the cultural economy: a case study of Hoxton, London, Urban Stud-ies, 46, pp. 10411061.

Sassen, S. (1994) Cities in a World Economy. Thousand Oaks, CA: Pine Forge.

Scott, A. J. (1999) The US recorded music in-dustry: on the relations between organiza-tion, location, and creativity in the cultural economy, Environment and Planning A, 31, pp. 19651984.

Scott, A. J. (2000) The Cultural Economy of Cities. London: Sage.

Scott, A. J. (2004) Cultural-products industries and urban economic development prospects for growth and market contestation in global context, Urban Affairs Review, 39, pp. 461490.

Sereda, P. (2007) Culture employment in a North American context, 1981 to 2001. Research paper, Culture, Tourism and the Centre for Education Statistics, Statistics Canada, Ottawa, ON.

Shapiro, D., Abercrombie, N., Lash, S. and Lury, C. (1992) Flexible specialisation in the culture industries, in: H. Ernste and V. Meier (Eds)



Regional Development and Contemporary In-dustrial Response: Extending Flexible Speciali-sation, pp. 179194. London: Belhaven.

Statistics Canada (2004) Canadian framework for culture statistics. Culture Statistics Program, Culture, Tourism and the Centre for Educa-tion Statistics, Statistics Canada, Ottawa, ON (http://www.statcan.gc.ca/bsolc/olc-cel/olc-cel?catno=81-595-MIE2004021&lang=eng).

Statistics Canada (2005) The time it takes to get to work and back. General Social Survey on Time Use: Cycle 19, Social and Aboriginal Statistics Division, Statistics Canada, Ottawa, ON.

Storper, M. and Christopherson, S. (1987) Flex-ible specialization and regional industrial agglomeration: the case of the U.S. motion picture industry, Annals of the Association of American Geographers, 77, pp. 104117.

Waitt, G. (2006) Creative small cities: city-scapes, power and the arts, in: D. Bell and

M. Jayne (Eds) Small Cities: Urban Experiences beyond the Metropolis, pp. 169183. New York: Routledge.

Waitt, G. and Gibson, C. (2009) Creative small cities: rethinking the creative economy in place, Urban Studies, 46, pp. 12231246.

Wenting, R. (2008) Spinoff dynamics and the spatial formation of the fashion design indus-try, 18582005, Journal of Economic Geogra-phy, 8, pp. 593614.

Wilkinson, P. F. and Murray, A. L. (1991) Centre and periphery: the impacts of the leisure in-dustry on a small town (Collingwood, Ontario), Society and Leisure, 14, pp. 235260.

Wojan, T. (2006) The emergence of rural artistic havens: a first look. Paper presented to the An-nual Meetings of the Southern Regional Science Association, St Augustine, Florida.

Zukin, S. (1995) The Cultures of Cities. Malden, MA: Blackwell Publishing.


Documents

Urban Stud 2012 Denis Jacob 97 114