micropolitan location of the software & videogames ...5 data source from institut català de les...
Post on 09-Feb-2020
0 Views
Preview:
TRANSCRIPT
1
Micropolitan Location of the Software & Videogames
Industry: an insight on the case of Barcelona*
Autores y e-mail de la persona de contacto: Carles Méndez-Ortega: carles.mendez@estudiants.urv.cat Josep-Maria Arauzo-Carod: josepmaria.arauzo@urv.cat Departamento: Departament d’Economia Universidad: Universidad Rovira i Virgili Área Temática: Indústrias Creativas Resumen:
This paper analyses location patters of Software and Videogames industries at a
micropolitan scale for the metropolitan area of Barcelona. These are key industries in
developed economies that benefits from agglomeration economies, skilled labour and,
generally speaking, spillover effects, that tends to cluster at bigger metropolitan areas,
but less is known about its detailed location patterns inside these areas. We contribute
by identifying how Software and Videogames industries firms concentrate in certain
areas of the metropolitan area. Our empirical application includes using Nearest
Neighbour Index (NNI) and M-functions, as well as local spatial autocorrelation
indicators.
Palabras Clave: R12, C60, L86 Clasificación JEL: Software Industry, Videogames Industry, micropolitan analysis, spatial location patterns
*This article has benefited from funding from ECO2013-41310-R, ECO2014-55553-P, the “Xarxa de Referència d’R+D+I en Economia i Polítiques Públiques” and research program SGR (2014 SGR 299). We would like to acknowledge comments received at INFER Annual Conference 2016, and also suggestions by E. Coll-Martínez. Any errors are, of course, our own.
2
1. Motivation
Software industry has transformed manner organization, businesses, and employees
coordinate and work. The impact of this industry on global economy can be measured
by the increase in technical progress, enhanced productivity and innovations.
Concretely, Software refers to computer programs or data (bits and bytes stored), which
can be stored electronically and it is used by the computer processor to perform several
tasks.
Software has gone from obscurity to indispensability in less than fifty years. Since the
middle of the 70’s, with the emergence of the Computer Revolution, it has led to the
appearance of this industry. In the 80’s, with the incorporation of the Personal
Computer (PC) and the developments in the field of Software and in the 90’s with the
development of the World Wide Web, this sector grew up exponentially. Nowadays
Software industry has a relevant importance for the current world economy,
representing a significant part of the information and communication technology sector,
which contributes 5.4% to global gross domestic product according to Dutta and Mia
(2010). Concretely in US are 1.87 million people working in this sector, earning one
hundred thousand dollars in average (more than the double of national average) and
represents more than 412 billion dollars of gross output for the US economy.1 Related to
Spain, this sector provides 360 thousand employees, earning more than 30 thousand
euros in average per year and employee and represents 30 billion euros of gross output
in the Spanish economy.2 This industry includes Software management, programming,
editing electronics, mobile Software (or apps) and Videogames industry, industry in
which we focus below.
Now we will focus in Videogames industry, a peculiar creative industry within Software
industry and halfway between Software industry and creative industries that shares
some characteristics with both industries (exponential growth, technical processes and
business features from Software industry and creative components from the creative
industries). It is important to notice that Videogames industry plays a key role as an
enhancer of economic activity, as a creative industry.
1OECD, STAN Dataset for Structural Analysis, ed. 2010. 2 INE (Instituto Nacional de Estadística), Sector ICT indicators, ed. 2013.
3
Videogames industry includes Software created for entertainment based on the
interaction between one or more persons and an electronic device that runs this game.
Devices include mobile phones, Personal Computer (PC), arcade machines or Video
Game Consoles among others. Videogames began in the early 1960’s in the United
States with the first Computer Game “Space Wars”, developed by Steve Russell at MIT
Lab. This game was inspired by science fiction and featured two spaceships shooting
beams to destroy each other. This videogame was never tried to be for sale; however, its
development occurred before the US government’s application of copyright laws to
computer Software. The first viable business in Videogames was in hands of Nolan
Bushnell, founder of Atari. He has been considered as “the first of that generation of
much-hyped super-successful high-tech entrepreneurs” (Aoyama & Izushi, 2003, pp.
427). This game initially created and later fuelled Videogames industry, focusing on its
beginnings in bigger markets, as those of Japan (with firms as Nintendo and Sega) and
the U.S. (with Atari at the beginning and Microsoft after).
Since the 1970’s, Videogames industry has moved from obscurity to a clear linkage to
the computer and digital revolution, giving the industry an exponential growth.
Nowadays, Videogames industry has a relevant importance for the current world
economy. Concretely, global spending of this industry was 70.8 billion dollars in 2014,
with a forecast of 89 billion dollars for 2018, presenting an annual global growth rate of
6.2% in mean.3 For Spain, this industry had about 1078 million dollars of incomes in
2013, with a forecast of 1238 million dollars for 2017 (annual growth rate is 3.3%).4 As
our analysis will focus in Barcelona, the income of this sector for Catalonia (region
located at north-east part of Spain whose capital is Barcelona) is 110 million dollars for
2014 and has more than 1600 workers. 5 In 2015, Catalan companies developed about
150 games, 40% for mobiles and 30% for online platforms.5
Location preferences for Software and Videogames industry are, mainly, inside big
metropolitan areas and, specifically, at the core of them, where they may easily found
young skilled professionals that are key for success of Software and Videogames
3 PWC: “Global entertainment and media outlook 2014-2018” 4 PWC España: “Global entertainment and media outlook 2013-2017” 5 Data source from Institut Català de les Empreses Culturals: “Empreses de videojocs a Catalunya 2015” for economic data and own estimations on the number of employees through our dataset.
4
activities. In this sense, there is empirical evidence showing the huge importance that
skilled labour has for these industries and, concretely, workers with upstanding
capabilities in terms of abstract skills, such as computer programming and Software
engineering (Autor et al., 2003). Availability of such skilled labour is a basic input for
high-tech firms (overrepresented in big urban areas) that even has a positive effect over
city growth, as shown by Berger and Frey (2015).
In view of previous concerns, we want to analyse case study of Software and
Videogames industry for a dynamic city, Barcelona, which has increasingly attracted
this type of activities in recent years, due to some potential factors, showed in the
number of university degrees offered in Barcelona universities dedicated to Videogames
design and Videogames management, and actual facts which proof the potentiality of
this city for this industry as the Barcelona Games World 2016, celebrated previously in
Madrid and now moved to Barcelona.6 Although Barcelona has a long-term tradition of
manufacturing activities, these were rapidly dismantled since the late seventies were
most of factories started to move from city centre, leaving enormous areas of free soil
available for further urban growth. Although some of these areas were transformed into
residential ones, there are some interesting experiences of urban renewal projects
aiming to transform them into high-tech districts. The most famous of them is the case
of Poblenou, a district quite close to city centre that in 2000 benefited from a renewal
plan form Barcelona City Council aiming to transform his area from a mature textile
manufacturing cluster to a high-tech cluster known as 22@, a successful policy that
attracted a large number of high-tech firms and knowledge based activities Viladecans-
Marsal and Arauzo-Carod (2012). In spite of success of 22@ policy not all new high-
tech related activities have decided to locate inside its borders. On the contrary, there
are many other core areas that have attracted these new firms, as shown by data
provided by Catalan Government (Institut Català de les Empreses Culturals) that covers
Videogames firms located in Barcelona.
According to previous theoretical concerns, to characteristics of Videogames and
Software industries and to the specificities of the Barcelona case, with this paper we aim
6 Universities as Universitat de Barcelona (UB), Universitat Politècnica de Catalunya (UPC) and Universitat Pompeu Fabra (UPF) presents Bachelor’s and Master degrees focus in Videogames topics.
5
to analyse location patterns of Software and Videogames firms inside metropolitan area
of Barcelona. In order to do that we have use some spatial oriented techniques as
Nearest Neighbour Index (NNI), Marcon functions (M-functions) and spatial
autocorrelation indicators (both global and local). Our preliminary results indicate that
Videogames firms tend to cluster around some subcentres inside Metropolitan area of
Barcelona (hereafter MAB) while Software firms are located within urban areas. Next
step will be to try to understand what explains these patterns and drives their location
determinants.
The results of this paper are counterpoised regard to Videogames industry if we
compare to the results obtained by the paper mentioned above of Viladecans-Marsal &
Arauzo-Carod (2012) which shows the success of the @22 district as a knowledge-
based cluster, we found a slight agglomeration pattern in this district for Videogames
firms.
The rest of the paper is organised as follows. Section 2 reviews the theoretical and
empirical literature about firms’ location determinants and recent contributions at a
micropolitan scale. Section 3 describes data and methodology. Section 4 introduces
some descriptive statistics and discusses results and, finally, section 5 presents main
conclusions.
2. Related literature
This paper is placed between two strands on economic literature closely linked,
industrial location determinants and agglomeration economies. The former, industrial
location determinants, is a recurrent topic in economics since the seminal work of
Alfred Marshall (1890) about industrial districts and Edgar and Hoover (1936), among
main contributors. Location literature exploring location decisions of new firms has
extensively growth in recent years, widening methodologies, spatial aggregation levels
and industries being analysed. That success is partially explained in terms of potential
utility of these contributions when designing public policies aiming to attract new
firms.7 The later, agglomeration economies, shows that economic activity has a strong
tendency to agglomerate, no matter industries, geographical areas or institutional 7 An extensive review of this literature can be found at Arauzo-Carod et al. (2010)
6
settings, albeit these characteristics modulate the degree of agglomeration effectively
reached. If we consider both strands together, we may conclude that i) firms (as well as
many other institutions) show a clear tendency to agglomerate around other firms and ii)
firms share specific location patterns (i.e., internal and external characteristics that guide
their location decisions) with similar firms.
Nevertheless, as in this paper we are mainly interested in spatial distribution of firms
rather than in whether this distribution comes from entry decisions or from firms
already existent that decide to agglomerate closer to other firms, we will concentrate our
analysis on agglomeration processes.
As contributions at agglomeration economics literature are quite heterogeneous and
approach location of economic activities in a quite different ways, there are several
attempts to organise and structure them. In this sense, Albert et al. (2012, p.109)
consider that “this literature has been influenced by two very different traditions,
economic geography and spatial statistics, and therefore has followed two different
paths”.
The first ones (economic geography) dealt with polygons (i.e., administrative areas) and
typically used measures as Gini or Herfindhal indices and did not take into account
space in their analyses and, consequently, may incur in important bias. Later, some
authors tried to overcome previous shortcomings by controlling for space,
agglomeration and concentration through several indices that become extremely popular
in this literature, as in Ellison and Glaeser (1997), Maurel and Sédillot (1999),
Rosenthal and Strange (2003) and Devereux et al. (2004). Finally, Duranton and
Overman (2005) assumed space as continuous, which allowed overcoming previous
limitations caused by MAUP like comparing results across different spatial aggregation
levels. The second ones (spatial statistics) dealt with points (i.e., firms), considering
only closest points (e.g., Nearest Neighbour Index) or taking into account all neighbours
instead of only the closest (e.g., K functions). Concretely, K functions have been
initially developed by Ripley (1977) and later improved by Marcon and Puech, (2012),
this type of measures were tested by Arbia (2001) which shows incentives to use
continuous space in these types of analysis using a continuous space model.
7
With the introduction of these tools, numerous papers were published analysing spatial
patterns of particular industries, such as the food stores in Italy analysed by Arbia et al.
(2015), media industries in London by Karlsson and Picard (2011), Espa et al. (2013)
with high tech industries in Milan, manufacturing firms in Paris done by Guillain and le
Gallo (2010) and for all Europe by Brülhart (2001) or in the case of the tree location in
U.S. forests, in an ecological and environmental framework by Li and Zhang (2007)
among others.
Focus in agglomeration processes related to the cities, an earlier work of Glaeser (1998)
analysed the positive and negative effects of cities and how local governments can solve
these arising problems; as well Anderson and Bogart (2001) and Coffey and Shearmur
(2002) analyse characteristics of the employment centres and compare it with the size
of the metropolitan areas and how the employment on cities is changing, appearing a
decentralization process of the employment in relative terms from the Central Business
District (CBD) to another parts of the city and metropolitan area, but not in absolute
terms, creating a polycentric structure and not a generalized dispersion. In Berger and
Frey (2015) mentioned above, it shows the relevant role that city plays on the research
of abstract skill workers, and how this affects positively to the location and profits for
the firms and growth of the cities that are intensive in this type of workers. Also Currid
(2006) in the case of New York city, analyses role of the city as a bastion of creativity,
culture and artist production, which attracts to creative firms and how this affects
positively to growth of the cities.
Apart from the general framework of this paper in terms of location of economic
activity and agglomeration economies, this paper covers specifically Videogames
industry, an area for which analyses are still pretty scarce, both because it is a young
industry and because it is between Software industry and creative industries (Lorenzen
and Frederiksen, 2008), that’s why we introduce main findings related with location
determinants of all these areas: Software industries, creative industries and Videogames
industry.
Firstly, regarding location analyses of Software industries, there are a widely literature
related to the location determinants across countries (e.g., Parthasarathy , 2004, for
India and Ó Riain, 1997, for Ireland). It’s a fact that this is an industry closely related to
8
the networking and spillovers effects, which plays a relevant role on the location of
these firms. There are some studies that proof the benefits of networking as in Cassiman
and Veugelers (2002), which analyse the positive effect of the spillovers due to R&D
cooperation and how the external information flows affects positively to firms, to the
agency industry (linked to the computer industries in terms of information and
connections across firms), Arzaghi and Henderson (2008) show the positive effect of be
located in a same area in terms of productivity because the information plays a relevant
role. Furthermore, Morgan (2004) explains how the location of this type of firms matter,
since distance can destroy the capacity of communication technologies and information
where social intensity is combined with spatial range, also the fact that physical
proximity across these firms can be essential for some forms of knowledge exchange
and additionally charting the growth of innovation systems on the territory.
Secondly, empirical analyses about creative industries cover a wide range of activities,
both from manufacturing and services. As it is happening with computer industry, the
location determinants and clustering of this industry it is widely discussed at all levels,
as continental level (Boix et al., 2015, for European regions), for specific countries
(Cruz and Teixeira, 2014 for Portugal and Lazzeretti et al., 2012, for Spain and Italy),
and also at city (Catungal et al., 2009, for the Canadian city of Toronto) and
metropolitan levels (Coll-Martínez et al., 2016, for the Metropolitan Area of Barcelona).
All these papers share some common characteristics, showing that creative firms are
highly clustered, mostly localized in metropolitan areas, around medium and large
cities, cross borders and areas with huge concentration of creative and knowledge-based
activities, places with high tolerance and open environments and higher education at a
regional level, among others.
Thirdly, as stated before specific analyses for Videogames industries are quite scarce,
partially because these activities have been traditionally considered jointly with other
computer-related activities such as Software design. Additionally, these contributions
are quite heterogeneous and cover different areas about location determinants. In this
sense, whilst Johns (2006) focus on the importance of the networks in this industry in
order to demonstrate the utility of Global Production Network approaches and the
uneven impact on the globalization processes, other authors try to explain intraurban
determinants of computer games industry alongside with other types of creative firms
9
(e.g., Murphy et al., 2015, for the specific case of Dublin) or to carry a broader
geographical approach in which clusters of different size are considered (e.g., Tschang
and Vang, 2008, for the U.S.).
3. Data and methodology
3.1 Data
Data used in this paper comes from two sources, Àrea de Cultura Digital (Institut
Català de les Empreses Culturals) and SABI (Sistema de Análisis de Balances
Ibéricos). The first one is a public institution from Catalan Government charged of
digital media activities like Videogames, books and music, among others; and the
second one uses data from the Mercantile Register. The data obtained for Software,
Videogames and editing electronics firms were from SABI Database, whilst
Videogames firms data were obtained from Àrea de Cultura Digital, which provides an
extensive list of Videogames firms obtained through day-to-day interaction of this
institution with firms operating in this industry, SABI allows to complete individual
data for each Videogames firm as number of employees, sales and assets, among other
characteristics while simultaneously it allows to obtain Software firms dataset.
Accordingly, for Videogames firms, we have carried out a matching between both
sources and, in some cases, we have contacted firms directly if firms listed in Àrea de
Cultura Digital were not compiled by SABI. Our final dataset is composed by 104
Videogames firms in Catalonia (25.15% of the Spanish total in 2015), although some of
them (17) do not have enough data to be georeferenced.8
As we indicate at the introductory section we focus on Metropolitan Area of Barcelona
(MAB, hereafter), located in Catalonia, northeast Spain. MAB (see Figure 1) covers an
area of 636 km2, has around 3,2 million inhabitants and includes 36 municipalities.9
8 Data provided from Generalitat de Catalunya. 9 Municipalities of MAB are Barcelona, l’Hospitalet de Llobregat, Badalona, Santa Coloma de Gramenet, Cornellà de Llobregat, Sant Cugat del Vallès, Sant Boi de Llobregat, Viladecans, el Prat de Llobregat, Castelldefels, Cerdanyola del Vallès, Esplugues de Llobregat, Gavà, Sant Feliu de Llobregat, Ripollet, Montcada i Reixac, Sant Adrià de Besòs, Sant Joan Despí, Barberà del Vallès, Sant Vicenç dels Horts, Sant Andreu de la Barca, Molins de Rei, Sant Just Desvern, Corbera de Llobregat, Badia del Vallès, Castellbisbal, Pallejà, Montgat, Cervelló, Tiana, Santa Coloma de Cervelló, Begues, Torrelles de Llobregat, el Papiol, Sant Climent de Llobregat, and la Palma de Cervelló. See http://www.amb.cat/ for details.
10
Figure 1: Metropolitan Area of Barcelona (MAB)
Source: http://www.amb.cat/
The main reason that makes us concentrate and only analyse the firms situated on the
MAB is the highest number of firms situated around Barcelona. The number of these
firms From the 104 Videogames firms from our data set, 74 are located in the MAB (54
of them directly located in Barcelona city); capturing with our analysis more than 71%
of Videogames industry location patterns in Catalonia, also related to Software,
Videogames and editing electronics firms (hereafter SVE), there is a high number of
firms located in the MAB, 858 SVE firms including Videogames firms (see Figure 2
and 3 for Point maps and Figure 4 and 5 for Heat maps overview). Regard to our
dataset, 2 Videogames firms do not have physical headquarters and we have 15
Videogames firms with an unknown location (4 of them located in the MAB, but it was
impossible to obtain their location).
11
Figure 2: Firms from SVE industry in the MAB
Source: SABI, Institut Català de les Empreses Culturals and own elaboration. Figure 3: Firms from Videogames industry in the MAB
Source: SABI, Institut Català de les Empreses Culturals and own elaboration.
12
Figure 4: Heat map from SVE industry in the MAB
Source: SABI, Institut Català de les Empreses Culturals and own elaboration. Figure 5: Heat map from Videogames industry in the MAB
Source: SABI, Institut Català de les Empreses Culturals and own elaboration.
13
Regard to the creative industries dataset, we used data from SABI using the
classification from Boix and Lazzeretti (2012) where they summarize several
classifications of creative industries made by cultural agencies and international expert
groups (i.e. OECD -Organisation for Economic Co-operation and Development , WIPO
– World International property Organization or the UNCTAD – United Nations
Conference on Trade and Development among others). According to this classification
we obtain 17 categories related to creative industries (see Table 1).
Table 1: Firms of Creative industries in the MAB
Industry/Classification
Acronym
Nº of firms
Share %
Advertising and related Services ADV 1218 15.59
Architecture and engineering AE 1587 20.31
Art and antiques trade ART 47 0.60
Crafts CR 49 0.63
Edition ED 538 6.89
Fashion FA 380 4.86
Graphic Arts GA 927 11.87
Heritage, Cultural sites and recreational Services HE 691 8.85
Investigation and development creative IDC 148 1.89
Jewellery, musical instruments, toys and games JEW 105 1.34
Music and music studies MU 74 0.95
Performing Arts PA 264 3.38
Photograph PHO 161 2.06
Radio and TV RTV 66 0.84
Software, Videogames and editing electronics SVE 858* 10.98
Specialized services design SSD 272 3.48
Video and film industries VFI 427 5.47
Total number of firms 7812 100.00 Source: SABI. (*) 4 firms from SVE are not georeferenced.
As we see in the table above, there are a total of 7812 creative firms in the MAB,
highlighting the high share on the number of firms with architecture and engineering
activities (more than 20% of total creative firms). In order to obtain a more reliable
creative firms dataset, firms whose physical address is not available, or not located
inside the MAB, or without data regarding number of employees, or extinguished are
not included in this dataset.10 From this classification we obtain the Category of SVE
10 Initially we obtain from the SABI Database 19229 firms (from Barcelona province, at NUTS 3 level), after the filter we have discarded 11491 firms between missing data and firms not located in the MAB. To
14
firms, used in our analysis. Finally, in order to be plainer, in the next table there is a
firm’s summary sorting by firms with missing data and with all the data (see table 2).11
Table 2: Number of firms’ summary
Videogames firms (104) SVE firms (858) With all data With missing data With all data With missing data
Catalonia 87 17 No data* No data*
MAB 70 4 854 4
Barcelona 54 0 640 0
Source: Own elaboration. (*) We do not obtain the number of SVE firms for all Catalonia inasmuch as our analysis is focused in the MAB.
From each firm we will use the following data: location, age and number of employees.
3.2 Spatial autocorrelation (Moran’s I and LISA)
Next step is to focus in the spatial autocorrelation patterns in order to check whether
geographical position each firm is independent from location of the rest of firms
included in our dataset. As figure 2 suggests that there is some kind of spatial
dependence, we will analyse whether distribution of these firms is spatially
autocorrelated at urban level.
In order to calculate spatial autocorrelation we do need a spatial-neighbour matrix (W)
for the neighbouring criterion definition. In order to solve some data constraints we will
limit this analysis to firms located in Barcelona. Concretely, we will use
neighbourhoods of Barcelona (73) as geographical units. This W matrix can be
approached in several ways (k-nearest neighbours, contiguous neighbours, distance-
based neighbours among others), but considering that this is an intraurban analysis in
which districts are quite close to each other in a delimited space (Barcelona) we
considered that the most appropriate measure is contiguity; that is, two districts are
neighbours if they share a common border.
this dataset (7812) were included Videogames firms situated in the MAB (74) to Software, Videogames and editing electronics category, deleting duplicated firms. 11 To see the evolution of the SVE industry in the MAB and inside the city of Barcelona, this can be found in the Annex to this work (A1. and A2).
15
Once W is identified, we calculate two measures of spatial autocorrelation: Moran’s I
(Moran, 1948) and the Local Index of Spatial Association (hereafter LISA). Values of
Moran’s Index indicate a negative spatial autocorrelation if they range from -1 to 0; a
positive spatial autocorrelation if they range between 0 and 1; and a random distribution
if they are around 0.
3.3 Nearest Neighbour Index (NNI)
The Nearest-Neighbour Index (NNI) is an indicator that compares mean of the observed
distance between each point (e.g., firms in this case) and its nearest neighbour with the
expected mean distance, assuming a random distribution. NNI is formulated as follows:
Where is the mean observed nearest neighbour distance, is the total number of
points and is the total area. The values of NNI are interpreted as follows: 0 shows a
clustered pattern, value 1 presents a randomly distribution of the points and values
higher to 2 shows a regularly dispersed pattern or uniform pattern. As this is a measure
widely used, there are numerous papers that uses, as an example, Rehák (2012) uses the
NNI among others in order to explain the level of spatial concentration of firms and
identify spatial clusters of firms in creative industries in Slovakia.
3.4 M-functions
M-functions allow capturing spatial clustering and are attracting attention of researchers
in a growing way. Nevertheless, there are some previous measures that deserve our
attention.
In order to understand well the M-function, it is important previously to highlight the
Ripley’s K. This function developed by Ripley (1976) computes density of any set of
points (i.e., firms) for a particular radius of distance in order to control if there is
clustering at certain distances. Hence, distance radius are calculated around every point,
16
being that K(d) is the mean density of points for every distance radius (d) and this
density is divided by the mean density of points (n) in the whole area (a): n/a. At first
glance interpretation of Ripley’s K seems to be pretty similar to that of NNI, as
expected random values are compared with observed values. Concretely, if observed
values exceed expected values, there is a clustered pattern; and if observed values are
lower than expected ones, there is a dispersed pattern.
Later, M-functions were introduced by Marcon and Puech (2010) as a cumulative
function that gives the relative frequency of neighbours of a chosen type (denoted as )
up to every distance, compared to the same ratio in the whole study subject area. The
M-function is defined as follows:
Where is the total weight of points . With this function we can measure the
concentration of a certain industry around a given industry. It is noteworthy that in our
empirical application the variable will be the number of employees for each firm.
There are some examples of papers using the M-functions as a tool to explain the
agglomeration or dispersion patterns between industries, as an example, Moreno-
Monroy and Cruz (2015) use this distance-based methods for measure the
agglomeration patterns of formal and informal manufacturing activity within the
metropolitan area of Cali. Similarly, Coll-Martínez et al. (2016) analyse spatial patterns
of agglomeration of creative industries in the Metropolitan Area of Barcelona using M
and m functions. Values of these functions are represented in a plot indicating for each
distance an M-value, when this M-value is outside the confidence interval, indicates that
there is a presence of agglomeration (M-value in the upper part of the plot) or presence
of dispersion (M-value in the lower part of the plot) for each distance. When these
values are inside the confidence interval, the results are inconclusive (not significant).
17
4. Results
4.1 Some descriptive statistics
In this section we will present some descriptive statistics about the SVE and
Videogames industry and we will compare them with the whole dataset of creative
firms.
Table 3: Statistical descriptive information
Mean
Standard Deviation
Max Min Rank Median
Videogames Industry Workers 21.3000 59.4680 400 1 399 5 Age 5.6857 4.7504 22 0 22 4 Number of firms 74 SVE Industry Workers 19.1405 75.5122 1329 1 1328 4 Age 13.0012 7.6389 62 0 62 12 Number of firms 858 Creative Industry Workers 10.8266 58.8732 2480 1 2479 2 Age 16.0288 9.6332 144 0 144 14 Number of firms 7812 Source: SABI, Institut Català de les Empreses Culturals and own elaboration.
Table 3 presents a set of descriptive statistics for SVE and Videogames industry and the
rest of creative industries situated in the MAB. Concretely, we include 74 firms with all
the data belonging to Videogames industry, 858 belonging to SVE industry and 7,812
firms from creative industries.12
Descriptive statistics show that Videogames industry is a young industry (5.686 years in
average) compared with the SVE industry (13.001) and whole creative industries
(16.029). Videogames industry also includes bigger firms (21.300 employees in
average) than SVE industry (19.141) and whole creative industries (10.827).
Nevertheless, this difference is mainly due to a couple of Videogames firms larger than
200 employees. In any case, the maximum number of employees in Videogames
industry is 400 (for SVE is 1329), whilst in the creative industries is 2.480 employees.
12 These firms (Videogames and SVE Firms) are obtained from firm’s data with age and number of workers available, firms with missing in these values were not included in the analysis.
18
On the contrary, the minimum values are the same for all industries (1 employee). In
terms of age, the oldest Videogames firm is 22 years old and the oldest SVE firm is 62
years old while for all the creative firms, the oldest firm is from XIX century (144 years
old). The median values for Videogames industry are 4 for both variables (years and
workers), for SVE industry are 4 and 12 and whilst for the creative industries are 2 and
14 respectively.
Previous data shows that Videogames industry is a young industry that has a huge
potential growth.
4.2 Agglomeration of Software & Videogames industry
Nearest Neighbour Index
Table 4 shows data for the Nearest Neighbour Index (hereafter NNI) as well as
additional spatial information.
First and second column of Table 4 use all firms of the sample (854 SVE firms,
including Videogames firms and in the second column 70 Videogames firms, all located
in the MAB) and then the NNI values when dividing Videogames firms sample in terms
in age (smaller/younger than 6 years old) and size (bigger/smaller than 3 employees).13
In terms of whole number of firms, our results show a clearly agglomeration pattern for
both industries with a NNI value of 0.3342 for all SVE firms and 0.4766 for
Videogames firms (this indicates that SVE firms are more concentrated than
Videogames firms). Let’s say, given that number of Videogames firms and the spatial
area of the MAB whilst the expected (random) distance among closest firms is 0.015,
real distance is only 0.007.
13 The cut point of firm’s age was obtained by median whereas that in the size case, we considered that firms with more than 3 employees are not small firms for this type of industry (the median for the size is in 5 employees).
19
Videogames firms Age Videogames firms Size
All SVE firms Videogames firms
Old firms (more than 6 years)
New firms (5 years or less)
Big firms (more than 3 workers)
Small firms (3 workers or less)
Average observed distance 0.00166 0.00730 0.01306 0.00896 0.01921 0.01005
Average expected distance 0.00496 0.01532 0.02071 0.01901 0.02274 0.01846
Nearest neighbour index 0.33428 0.47660 0.63081 0.47140 0.84070 0.54472
Number of observations 854 70 34 36 25 45
Z-Value -37.21748 -8.37739 -4.11767 -6.06745 -1.52290 -5.84268 Source: Own elaboration
Table 4: Nearest Neighbour Index from SVE & Videogames Industry in the MAB.
20
Additionally, when we sort by Videogames firm age we get an interesting result,
showing that older firms (0.6308) are less agglomerated that newer ones (0.4714). That
result may indicate a different behaviour from new firms that tend to be located closer
to each other than older firms. This was an expected result, inasmuch as these new firms
are looking for places where the information and knowledge are shared and more open
environments (firm’s incubators, and technological research centres, among others) in
order to obtain a spillover effects.
Finally, we sort by Videogames firm size and we obtain as well an expected result, as
smaller firms (up to 3 employees) show a higher agglomeration pattern (0.5447) than
the bigger ones (0.8407). These results indicate that small firms are looking for places
closer to each other than bigger firms do. In line with age results, this is an expected
effect, as these firms will choose open areas and sites where share is easier, in order to
benefit from spillover effects and knowledge from other Videogames firms.
It is important to mention the link between size and age, as young firms are smaller than
old ones. Therefore, our results reveal that in Videogames industry, small and young
firms will be more concentrated and allocated closer each other than bigger and older
firms.
Moran’s I and LISA
In this section we calculate both global (Moran’s I) and local (LISA) measures of spatial
autocorrelation for SVE firms and Videogames firms. It is important to notice the
limitation of our analysis as we focus only in Videogames and SVE firms located inside
the city of Barcelona, without considering those located in the rest of the MAB (next
step is to include them).14 In order to calculate spatial autocorrelation we divide the city
of Barcelona in 73 neighbourhoods, as these are the main administrative units used by
city council. Concretely, there are 640 SVE firms (including Videogames firms) located
in Barcelona city of 54 are Videogames firms, as shown in Table 5, most of them
clustered at Eixample and Sant Martí districts.
14 We restrict the spatial scope due to some data constraints.
21
Table 5: Districts, Neighbourhoods (No.) and number of firms by Neighbourhood.
Source: Own elaboration. Note: The neighbourhood names list is in the Annex (Table A3).
District No. No. SVE firms
No. Videogames firms District No.
No. SVE firms
No. Videogames firms
District No. No. SVE firms
No. Videogames firms
Ciutat Vella 1 4 1 26 62 4 51 0 0
2 8 0 27 9 1 52 2 0
3 0 0 Gràcia 28 7 0 53 0 0
4 6 0 29 1 0 54 0 0
Eixample 5 8 0 30 6 0 55 0 0
6 12 0 31 25 2 56 0 0
7 107 8 32 10 1 Sant Andreu 57 0 0
8 52 4 Horta-Guinardó 33 12 0 58 0 0
9 32 3 34 1 0 59 1 0
10 13 0 35 2 0 60 3 1
Sants-Montjuïc 11 1 0 36 2 0 61 10 3
12 3 0 37 0 0 62 2 1
13 2 0 38 1 0 63 2 0
14 2 0 39 0 0 Sant Martí 64 6 1
15 2 0 40 1 0 65 4 0
16 2 0 41 7 1 66 37 4
17 4 0 42 0 0 67 10 1
18 6 1 43 3 1 68 15 1
Les Corts 19 43 4 Nou Barris 44 2 1 69 16 2
20 7 0 45 3 0 70 1 0
21 10 0 46 1 0 71 15 5
Sarrià-Sant Gervasi 22 0 0 47 0 0 72 3 0
23 10 0 48 7 2 73 0 0
24 6 0 49 0 0
25 11 1 50 0 0 TOTAL 640 54
22
Figure 6 shows local spatial autocorrelation (LISA) for the SVE firms at city-
neighbourhood level.15 This local approach highlights the high-high local spatial
autocorrelation existent in the Eixample district and surroundings, indicating the high
number of SVE firms located in these areas. The Ciutat Vella district presents a low-
high local spatial autocorrelation, indicating the small number of SVE firms located
there (compared with neighbouring districts). Is interesting to notice the low-low local
spatial autocorrelation existent in the north part of Nou Barris and Horta Guinardó
districts, indicating the negative local spatial autocorrelation of this part of Barcelona
city (places with low number of SVE firms compared to their neighbouring districts).
Figure 6: LISA for Barcelona city neighbourhoods (SVE firms)
Source: Own elaboration.
Figure 7 shows local spatial autocorrelation (LISA) for Videogames firms at city-
neighbourhood level.16 This local approach highlights the high-high local spatial
autocorrelation existent in the Eixample district (concretely the neigbourhoods of
Antiga Esquerra de l’Eixample, la Dreta de l’Eixample) the neigbourhoods of Sant
Gervasi–Galvany, Vila de Gràcia, el Camp d'en Grassot i Gràcia Nova and also
Poblenou neighbourhood, this implies that the number of Videogames firms in these
15 We have used a 0.05 significance level and 499 permutations. 16 We have used a 0.05 significance level and 499 permutations.
23
neighborhoods is high when compared with the number of Videogames firms in
adjacent neighborhoods. The neighbourhoods of el Camp de l'Arpa del Clot and la
Guineueta have a high-low local spatial autocorrelation an as before, this result
indicates that neighbourhoods have a large number of firms surrounded by other where
the number of firms is small. Results indicate also significant values for el Barri Gòtic,
Sant Pere, Santa Caterina I la Ribera , el Fort Pienc, la Sagrada Família, les Tres Torres,
Sant Antoni and el Camp de l'Arpa del Clot, where there is a low-high effect (i.e.,
neighbourhoods with a low number a firms surrounded by adjacent districts with a large
number of firms).
Figure 7: LISA for Barcelona city neighbourhoods (Videogames firms)
Source: Own elaboration.
In terms of the global measure of spatial autocorrelation (Moran’s I), for the SVE firms
case main results show a positive value (0.3617) indicating a positive spatial
autocorrelation for these firms at neighbourhood level in Barcelona, this also implies an
agglomeration effect of this industry at neighbourhood level in the city, indicating that
these firms are not random distributed or dispersed within Barcelona. Related to
Videogames firms, main results show a pattern close to a random distribution but
positive (0.1680). In any case, these are preliminary results that deserve additional work
to be done.
24
M-functions
In order to do an extensive analysis of the Videogames industry agglomeration patterns,
we show the intra-industry effect of Videogames and SVE industry and a set of inter-
industry effects. In order to carry out the analysis we will compare Videogames
industries with those of the creative industries. Additionally, we will go deeper into
some specific creative industries (Advertising and related services, Edition, Graphic
Arts, Software, Videogames and editing electronics and Video and Film industries) that
we guess that may be collocated with Videogames industries17. Firstly we will analyse
intra-industry effects and secondly inter-industry effects.
Figure 8 and 9 portraits the graphs of the M-function related to the SVE and
Videogames intra-industry effect respectively. Concretely, for the SVE firms, M-
function indicates that these firms present a high concentration (high M-values). Related
to Videogames firms, this graph also presents a high concentration until approximately
1000 meters radius. According to that evidence we may argue that SVE and
Videogames firms are agglomerated.
Figure 8: Intra-Industry agglomeration Software, Videogames and editing electronics industry
Source: Own elaboration.
17 All calculations are made at 0.05 significance level using 150 simulations.
25
Figure 9: Intra-Industry agglomeration Videogames industry
Source: Own elaboration.
Now we will analyse the inter-industry effects of the SVE firms with all the creative
firms, figures A4. and A5. show the M-functions in which SVE firms are compared
with all the creative firms. 18 The results indicate a statistically significant dispersed
effect in the first tram of the radius, indicating that SVE firms tend to be dispersed in
terms of respective partner and vice versa.
Next we will compare Videogames industry with all the creative industries and also
with some specific creative industries. Concretely, figures A6. and A7. show the M-
functions in which Videogames industry are compared with all the creative industries.
Results indicate a statistically significant dispersed effect in the first tram of the radius,
indicating that both industries tend to be dispersed in terms of respective partner.
Analysis of SVE industries and Advertising and related services is shown in figures A8.
and A9. For this pair of industries our results indicate that there is a slightly
agglomeration pattern of this industry in terms of respective pattern for values situated
18 All the figures related to inter-industry effects are in the Annex.
26
in the middle part of the radius (5000 meters approximately) as M-function values are
significant. Analysis of Videogames industries and Advertising and related services is
shown in figures A10. and A11. For this pair of industries our results indicate that there
are not agglomeration or dispersion patterns as M-function values are not significant.
Analysis of SVE industries and Edition is shown in figures A12. and A13. Results are
pretty similar to previous ones as there is no evidence of agglomeration (M-function
values are not significant). Analysis of Videogames industries and Edition is shown in
figures A14. and A15. The results are in line with previous ones: there is no evidence of
significant agglomeration.
Figures A16. and A17. show analysis of SVE industries and Graphic arts. The results
indicate a statistically significant dispersed effect in the first tram of the radius (until
6000 meters), indicating that both industries tend to be dispersed in terms of respective
partner and vice versa. Figures A18. and A19. show analyses of Videogames industries
and Graphic arts. The results are in line with the comparison of this industry with
Edition: there is no evidence of significant agglomeration.
Figures A20. and A21. show analysis of SVE industries and Video and Film industries.
For this pair of industries M-function indicates that there is evidence of agglomeration
in both ways (in the figure 26 this effect is slightly around 10000 meters of radius).
Figures A22. and A23. show analysis of Videogames industries and Video and Film
industries. For this pair of industries M-function indicates that there is evidence of
agglomeration Video and Film industries around Videogames industries, but only
slightly in the opposite way.
Finally, we analyse the case of Videogames industries and SVE without Videogames
firms, the result is shown in figures A24. and A25., being the results there is no
evidence of significant agglomeration between SVE and Videogames firms.
To conclude, in the next table we present a summary of all the M-function results (see
table 6).
27
Table 6: Inter-Industry and Intra-industry effects summary.
Categories SVE Firms Videogames
Firms
All Creative Dispersion Dispersion
ADV Agglomeration Not significant
ED Not significant Not significant
GA Dispersion Not significant
VFI Agglomeration Agglomeration
SVE Agglomeration(*) Not significant
Videogames Not significant Agglomeration(*)
Source: Own elaboration. (*) These fields represent the Intra-industry effect. Note: ADV (Advertising and Related Services), ED (Edition), GA (Graphic Arts), VFI (Video and Film industries) and SVE (Software, Videogames, and editing electronics). In this table are summarized all the results obtained regard to the M-functions intra-
industry effects and inter-industry effects. We observed that both industries
(Videogames and SVE) are agglomerated themselves, whereas that both also are
dispersed when we compare them with all the creative firms. These results glimpse that
these type of firms apparently are not attracted by other creative firm when we treat
these as a whole. Focus in particular cases, we observe the existence of an
agglomeration pattern of both industries with Video and film industries, where these
industries are agglomerated each other. Additionally, in the case of SVE firms, we
observe a dispersion pattern of these firms with Graphic Arts, indicating that both
industries are located away each other and an agglomeration pattern of SVE firms with
Advertising and related services, suggesting that these industries are agglomerated each
other.
5. Conclusions
Consistent with the main tenets of recent empirical literature analysing location patterns
of specific industries at a very detailed geographical levels, this paper shows that SVE
and Videogames industries follow similar patterns of other service industries as they
28
tend to cluster around some areas of the Metropolitan Area of Barcelona as a whole, and
the Barcelona city centre in particular. We have focused these industries in view of its
strategic relevance for more developed economies and also because they use to cluster
in and around big urban areas (i.e., Metropolitan Area of Barcelona hosts around 67%
of Videogames firms operating in Catalonia) attired by creative know-how
environment, skilled workforce, educational institutions providing specialised training
programs and worldwide professional events as Barcelona Games World 2016.
However, very few is known about their specific location patterns and, specially, about
whether they tend to co-locate close other creative industries.
Our results show that firms characteristics partially help to explain their location
preferences, as smaller / younger behave in a different way than bigger / older.
Concretely, for Videogames industry small and young firms show a more concentrated
pattern. Additionally, we have tested whether location of firms for these industries is
spatially autocorrelated, both using global (Moran’s I) and local measures (LISA). In
this sense, Moran’s I indicates a moderate spatial correlation for SVE industries, whilst
for Videogames industry evidence is even smaller. Nevertheless, local measures are of
big interest as they highlight some similarities and differences for both industries. On
the one side, main similarities arise from high-high correlation patterns at central
neighbourhoods (mainly around Barcelona’s main axis, Diagonal avenue) and at 22@
neighbourhood, an area where Barcelona’s city council has promoted a cluster of high-
tech firms. One the other side, main differences are in term of low-low areas, as while
for SVE industry there is a big are at low-income districts located NE of the city, there
is no a similar phenomena for Videogames, for which there is no any low-low
significant area.
M-function results validate previous conclusions from spatial autocorrelation measures
and suggest further insights for both industries. Concretely, they allow contextualizing
SVE and Videogames industries in terms of creative industries location patterns. In this
sense, both industries are not clustered around the whole range of creative industries.
Going deeper into creative industries classification show that patterns differ between
SVE and Videogames industries. Namely, whilst for the former there is a significant
clusterization pattern with Advertising and related services and Video and film
industries and a dispersion pattern with Graphic arts; for the later there is neither
29
clusterization or dispersion pattern (except for Video and film industries). These results
indicate that inter-industry interactions of SVE and Videogames industries are quite
specific and exist only for some creative activities.
Our results have important implications for policies aiming to attract and encourage
local development of firms from SVE and Videogames industries, as these results
provide some key facts about locational preferences for firms in these industries.
Although both industries are concentrated in and around Metropolitan Area of
Barcelona, concentration of Videogames industry is noticeable in Barcelona city centre
and, specially, around two subcentres located at i) both sides of Diagonal avenue and ii)
22@ district. In this sense, it is important that policy makers take into account these
firms’ preferences in order to provide some specific services that could make Barcelona
even a more appealing destination for them. As Videogames industry is rapidly growing
worldwide, the challenge for local policy makers is to continue attiring new firms and
related activities that could enhance competitiveness of Barcelona’s cluster.
Obviously, this analysis suffers from some limitations which have to be solved in future
research. Firstly, as we focus on official definition of Metropolitan Area of Barcelona
we consider 36 municipalities, but there are alternative classifications of the same area
made up from a functional point of view that (e.g., 165 municipalities according to
Spanish Ministry of Development (Ministerio de Fomento)19. Secondly, as these are
very dynamic industries it is not clear at all whether all firms are being captured in our
datasets. Therefore, we aim to check whether our results still hold after taken into
account these issues.
We will leave for further research the analysis of whether our conclusions hold for
alternative definitions of the Metropolitan Area of Barcelona. In any case, our results
provide a clear picture of location pattern of SVE and Videogames industries using the
most standard geographical grouping around Barcelona. Additionally, spatial
autocorrelation will be also extended for the whole area.
19 “Las Grandes Áreas Urbanas y sus municipios 2015”. Source: http://www.fomento.gob.es/NR/ rdonlyres/416CE7FD-A6B0-431D-881B-D1F07664795E/133984/listado__2015_2.pdf
30
6. References
Albert, J. M., Casanova, M. R., & Orts, V. (2012). Spatial location patterns of Spanish manufacturing firms. Papers in Regional Science, 91(1), 107–136.
Anderson, N. B., & Bogart, W. T. (2001). The Structure of Sprawl: Identifying and Characterizing Employment Centers in Polycentric Metropolitan Areas. American Journal of Economics and Sociology, 60(1), 147–169.
Aoyama, Y., & Izushi, H. (2003). Hardware gimmick or cultural innovation? Technologial, cultural, and social foundations of Japanese video game industry. Research Policy, 32(3), 423–444.
Arauzo-Carod, J. M., Liviano-Solis, D., & Manjón-Antolín, M. (2010). Empirical Studies in Industrial Location: an Assessment of Their Methods and Results. Journal of Regional Science, 50(3), 685–711.
Arbia, G. (2001). Modelling the Geography of Economic Activities on a Continuous Space. Papers in Regional Science, 80(4), 411–424.
Arbia, G., Cella, P., Espa, G., & Giuliani, D. (2015). A micro spatial analysis of firm demography: the case of food stores in the area of Trento (Italy). Empirical Economics, 48(3), 923–937.
Arzaghi, M., & Henderson, J. V. (2008). Networking off Madison Avenue. Review of Economic Studies, 75(4), 1011–1038.
Autor, D. H., Levy, F., & Murnane, R. J. (2003). The Skill Content of Recent Technological Change: An Empirical Exploration. Quaterly Journal of Economics, 118(November).
Berger, T., & Frey, C. B. (2015). Did the Computer Revolution Shift the Fortunes of U.S. Cities ? Technology shocks and the geography of new jobs. Regional Science and Urban Economics, 57, 38–45.
Boix, R., Hervás-Oliver, J. L., & De Miguel-Molina, B. (2015). Micro-geographies of creative industries clusters in Europe: From hot spots to assemblages. Papers in Regional Science, 94(4), 753–772.
Boix, R., & Lazzeretti, L. (2012). Las industrias creativas en España: Una panorámica. Investigaciones Regionales, (22), 181–205.
Brülhart, M. (2001). Evolving geographical concentration of European manufacturing industries. Review of World Economics (Weltwirtschaftliches Archiv), 137(2), 215–243.
Cassiman, B., & Veugelers, R. (2002). R & D Cooperation and Spillovers: Some empirical evidence. The American Economic Review, 92(4), 1169–1184.
Catungal, J. P., Leslie, D., & Hii, Y. (2009). Geographies of Displacement in the Creative City: The Case of Liberty Village, Toronto. Urban Studies, 46(5-6), 1095–1114.
Coffey, W. J., & Shearmur, R. G. (2002). Agglomeration and Dispersion of High-order Service Employment in the Montreal Metropolitan Region, 1981–96. Urban Studies, 39(3), 359 – 378.
31
Coll-Martínez, E.; Moreno-Monroy, A.I. and Arauzo-Carod, J.M. (2016): Agglomeration of Creative Industries: an Intrametropolitan Analysis for Barcelona. Working Paper CREIP 16-2016.
Cruz, S. S., & Teixeira, A. A. C. (2014). The determinants of spatial location of creative industries start-ups: Evidence from Portugal using a discrete choice model approach. FEP Working Papers.
Currid, E. (2006). New York as a Global Creative Hub: A Competitive Analysis of Four Theories on World Cities. Economic Development Quarterly, 20(4), 330–350.
Devereux, M. P., Griffith, R., & Simpson, H. (2004). The geographic distribution of production activity in the UK. Regional Science and Urban Economics, 34, 533–564.
Duranton, G., & Overman, H. G. (2005). Testing for Localization Using Micro-Geographic Data. The Review of Economic Studies, 72(4), 1077–1106.
Dutta, S., & Mia, I. (2010). The global information technology report 2009 – 2010 ICT for Sustainability. Technology.
Edgar, M., & Hoover, J. (1936). The Measurement of Industrial Localization. The Review of Economics and Statistics, 18, 162–171.
Ellison, G., & Glaeser, E. L. (1997). Geographic Concentration in U.S. Manufacturing Industries: A Dartboard Approach. Journal of Political Economy, 105(5), 889.
Espa, G., Arbia, G., & Giuliani, D. (2013). Conditional versus unconditional industrial agglomeration: Disentangling spatial dependence and spatial heterogeneity in the analysis of ICT firms’ distribution in Milan. Journal of Geographical Systems, 15(1), 31–50.
Glaeser, E. L. (1998). Are Cities Dying ? The Journal of Economic Perspectives, 12(2), 139–160.
Guillain, R., & le Gallo, J. (2010). Agglomeration and dispersion of economic activities in and around Paris: An exploratory spatial data analysis. Environment and Planning B: Planning and Design, 37(6), 961–981.
Johns, J. (2006). Video games production networks: Value capture, power relations and embeddedness. Journal of Economic Geography, 6(2), 151–180.
Karlsson, C., & Picard, R. G. (2011). Microclustering of the media industries in London. In Media clusters: Spatial agglomeration and content capabilities (pp. 120–135).
Lazzeretti, L., Capone, F., & Boix, R. (2012). Reasons for Clustering of Creative Industries in Italy and Spain. European Planning Studies, 20(8), 1243–1262.
Li, F., & Zhang, L. (2007). Comparison of point pattern analysis methods for classifying the spatial distributions of spruce-fir stands in the north-east USA. Forestry, 80(3), 337–349.
Lorenzen, M., & Frederiksen, L. (2008). Why do cultural industries cluster? Localization, urbanization, products and projects. In P. Cooke & L. Lazzeretti (Eds.), Creative cities, cultural clusters and local economic development (pp. 155–179).
Marcon, E., & Puech, F. (2010). Measures of the geographic concentration of industries: Improving distance-based methods. Journal of Economic Geography, 10(5), 745–762.
32
Marcon, E., & Puech, F. (2012). A typology of distance-based measures of spatial concentration. Hal Shs, 00679993.
Marshall, A. (1890). Principles of Economics. Macmillan: London.
Maurel, F., & Sédillot, B. (1999). A measure of the geographic concentration in french manufacturing industries. Regional Science and Urban Economics, 29(5), 575–604.
Ministerio de Fomento. (2015). Las grandes areas urbanas y sus municipios 2015. Avaiable in: http://www.fomento.gob.es/NR/ rdonlyres/416CE7FD-A6B0-431D-881B-D1F07664795E/133984/listado__2015_2.pdf
Moran, P. A. P. (1948). The Interpretation of Statistical Maps. Journal of the Royal Statistical Society. Series B, 10(2), 243–251.
Moreno-Monroy, A. I., & Cruz, G. A. G. (2015). Intra-Metropolitan Agglomeration of Formal and Informal Manufacturing Activity: Evidence from Cali, Colombia. Tijdschrift Voor Economische En Sociale Geografie, 00(00).
Morgan, K. J. (2004). The Exaggerated Death of Geography: Learning, Proximity and Territorial Innovation Systems. Journal of Economic Geography, 4(1), 3–21.
Murphy, E., Fox-Rogers, L., & Redmond, D. (2015). Location Decision Making of “Creative” Industries: The Media and Computer Game Sectors in Dublin, Ireland. Growth and Change, 46(1), 97–113.
Ó Riain, S. (1997). An Offshore Silicon Valley? Competition & Change 2, 1(2), 175–212.
Parthasarathy, B. (2004). India’s Silicon Valley or Silicon Valley's India? Socially embedding the computer Software industry in Bangalore. International Journal of Urban and Regional Research, 28(3), 664–685.
Rehák, Š. (2012). Exploring Spatial Patterns of Creative Industries with Firm Level Micro Geographic Data. Skúmanie Priestorového Usporiadania Kreatívnych Odvetví S Využitím Mikro Geografických údajov Firiem Exploring, (January), 10–35.
Ripley, B. D. (1976). The Second-Order Analysis of Stationary Point Processes. Journal of Applied Probability, 13(2), 255–266.
Ripley, B. D. (1977). Modelling Spatial Patterns. Journal of the Royal Statistical Society . Series B, 39(2), 172–212.
Rosenthal, S. S., & Strange, W. C. (2003). Geography, Industrial Organization, and Agglomeration. The Review of Economics and Statistics, 85(May), 377–393.
Tschang, F. T., & Vang, J. (2008). Explaining the Spatial Organization of Creative Industries: the Case of the U.S. Videogames Industry. Enterpreneurship and Innovation - Organizations, Institutions, Systems and Regions.
Viladecans-Marsal, E., & Arauzo-Carod, J. M. (2012). Can a knowledge-based cluster be created? The case of the Barcelona 22at district. Papers in Regional Science, 91(2), 377–400.
33
7. ANNEX
A1. Evolution of the number of SVE firms in the MAB. Year 1995 1998 2000 2004 2008 2012 2015 No of firms 142 212 280 440 629 814 854* *Firms with all data (See table 2).
34
A2. Evolution of the number of SVE firms in the city of Barcelona.
Year 1995 1998 2000 2004 2008 2012 2015 No of firms 121 176 230 338 479 608 640* *Firms with all data (See table 2).
35
Table A3. Barcelona Neighborhood name list. No. Neighborhood No. Neighborhood No. Neighborhood
1 el Raval 26 Sant Gervasi - Galvany 51 Verdun
2 el Barri Gòtic 27 el Putxet i el Farró 52 la Prosperitat
3 la Barceloneta 28 Vallcarca i els Penitents 53 la Trinitat Nova
4 Sant Pere, Santa Caterina i la Ribera 29 el Coll 54 Torre Baró
5 el Fort Pienc 30 la Salut 55 Ciutat Meridiana
6 la Sagrada Família 31 la Vila de Gràcia 56 Vallbona
7 la Dreta de l'Eixample 32
el Camp d'en Grassot i Gràcia Nova 57 la Trinitat Vella
8 l'Antiga Esquerra de l'Eixample 33 el Baix Guinardó 58 Baró de Viver
9 la Nova Esquerra de l'Eixample 34 Can Baró 59 el Bon Pastor
10 Sant Antoni 35 el Guinardó 60 Sant Andreu
11 el Poble Sec 36 la Font d'en Fargues 61 la Sagrera
12 la Marina del Prat Vermell 37 el Carmel 62 el Congrés i els Indians
13 la Marina de Port 38 la Teixonera 63 Navas
14 la Font de la Guatlla 39 Sant Genís dels Agudells 64 el Camp de l'Arpa del Clot
15 Hostafrancs 40 Montbau 65 el Clot
16 la Bordeta 41 la Vall d'Hebron 66
el Parc i la Llacuna del Poblenou
17 Sants - Badal 42 la Clota 67 la Vila Olímpica del Poblenou
18 Sants 43 Horta 68 el Poblenou
19 les Corts 44
Vilapicina i la Torre Llobeta
69 Diagonal Mar i el Front Marítim del Poblenou
20 la Maternitat i Sant Ramon 45 Porta 70 el Besòs i el Maresme
21 Pedralbes 46 el Turó de la Peira 71 Provençals del Poblenou
22 Vallvidrera, el Tibidabo i les Planes 47 Can Peguera 72 Sant Martí de Provençals
23 Sarrià 48 la Guineueta 73 la Verneda i la Pau
24 les Tres Torres 49 Canyelles
25 Sant Gervasi - la Bonanova 50 les Roquetes
Source: Ajuntament de Barcelona.
36
Figure A4. Inter-Industry agglomeration Software, Videogames and editing electronics vs. Creative
Source: Own elaboration. Figure A5. Inter-Industry agglomeration Creative vs. Software, Videogames and
editing electronics
Source: Own elaboration.
37
Figure A6. Inter-Industry agglomeration Videogames vs. Creative
Source: Own elaboration. Figure A7. Inter-Industry agglomeration Creative vs. Videogames
38
Source: Own elaboration. Figure A8. Inter-Industry agglomeration Software, Videogames and editing electronics vs. Advertising and related services
Source: Own elaboration. Figure A9. Inter-Industry agglomeration Advertising and related services vs. Software, Videogames and editing electronics
39
Source: Own elaboration. Figure A10. Inter-Industry agglomeration Videogames vs. Advertising and related services
Source: Own elaboration. Figure A11. Inter-Industry agglomeration Advertising and related services vs. Videogames
Source: Own elaboration.
40
Figure A12. Inter-Industry agglomeration Software, Videogames and editing electronics vs. Edition
Source: Own elaboration. Figure A13. Inter-Industry agglomeration Edition vs. Software, Videogames and editing electronics
Source: Own elaboration.
41
Figure A14. Inter-Industry agglomeration Videogames vs. Edition
Source: Own elaboration.
Figure A15. Inter-Industry agglomeration Edition vs. Videogames
Source: Own elaboration.
42
Figure A16. Inter-Industry agglomeration Software, Videogames and editing electronics vs. Graphic Arts
Source: Own elaboration. Figure A17. Inter-Industry agglomeration. Graphic Arts vs. Software, Videogames and editing electronics
Source: Own elaboration.
43
Figure A18. Inter-Industry agglomeration Videogames vs. Graphic Arts
Source: Own elaboration. Figure A19. Inter-Industry agglomeration Graphic Arts vs. Videogames
Source: Own elaboration.
44
Figure A20. Inter-Industry agglomeration Software, Videogames and editing electronics vs. Video and Film industries
Source: Own elaboration. Figure A21. Inter-Industry agglomeration Video and Film industries vs. Software, Videogames and editing electronics
Source: Own elaboration.
45
Figure A22. Inter-Industry agglomeration Videogames vs. Video and Film industries
Source: Own elaboration. Figure A23. Inter-Industry agglomeration Video and Film industries vs. Videogames
Source: Own elaboration.
46
Figure A24. Inter-Industry agglomeration Videogames vs. Software, Videogames and editing electronics (Without Videogames firms)
Source: Own elaboration. Figure A25. Inter-Industry agglomeration Software, Videogames and editing electronics (Without Videogames firms) vs. Videogames
Source: Own elaboration.
top related