1

INTERNATIONAL MIGRATION AND DEMOGRAPHIC CHANGE

IN THE METROPOLITAN AREA OF BARCELONA: A DEMO-SPATIAL ANALYSIS AT THE CENSUS TRACT LEVEL

Juan GALEANO (jgaleano@ced.uab.es),

Centre d’Estudis Demogràfics

Jordi BAYONA-i-CARRASCO, (jbayona@ced.uab.es) Centre d’Estudis Demogràfics & Universitat de Barcelona

Abstract: During the first decade of the 21st Century the Metropolitan Region of Barcelona experienced an enormous demographic transformation linked to incorporation of more than 300 thousand foreign immigrants. The residential settlement of newcomers changed drastically the composition of the human landscape at all territorial levels, both national and local, imposing to policy makers and migration managers a great challenge at managing increasing population diversity. Within this context, two linked phenomena are perceived as particularly relevant for social cohesion: the degree of residential segregation between different population groups and the spatial concentration of these newcomers. In order to assess the degree and trend of spatial assimilation of different immigrant groups and the process of formation of concentration areas and its evolution over time, we combine analysis on residential segregation at the metropolitan level with the implementation of the local indicator of spatial association Moran’s, that allow us to locate clustering areas at the census tract level. Our results show the coexistence of a general trend towards spatial assimilation of the foreign-born populations at the regional level (except for Western Europeans), along with the consolidation, and even extension, of ethnic concentrations at the local one. Keywords: International Migration; Residential Segregation; Ethnic Concentration; Metropolitan Area of Barcelona.

2

INTERNATIONAL MIGRATION AND DEMOGRAPHIC CHANGE IN THE METROPOLITAN AREA OF BARCELONA:

A DEMO-SPATIAL ANALYSIS AT THE CENSUS TRACT LEVEL

1. INTRODUCTION1 The arrival of the century was accompanied by a deep transformation of human geography of the whole Spain, due to the incorporation of more than 5 million foreign immigrants between 2000 and 2010. Particularly, the migration inflow had a notorious impact in the region of Catalonia; this single autonomous community absorbed the 22% of the total inflow. In the case of the Metropolitan Area of Barcelona (MAB), the aforementioned transformation can be illustrated in demographic terms, and broadly speaking, by the contrary trend followed by the number of people who lived according to their birthplace. While the number of Spanish-born population decreased from 2.75 to 2.6 million between 2003 and 2013, due mainly to a process of intense sub-urbanization (Pujadas, 2009; López-Gay, 2014), the foreign-born population almost doubles (from 320 to 620 thousand) during the same period, rising its share up to 19.2% of the total population. Alongside with its magnitude and intensity, the great diversification of the countries of origin (Chacón-Rodríguez, 2002), and the rapid diffusion form the gateway cities (Bayona and Gil-Alonso, 2012; Sabater et al., 2012) were two of the most relevant features of the inflow. From a geo-demographic perspective, two linked phenomena are perceived as particularly relevant for social cohesion: the degree of residential segregation between immigrants and autochthonous, and the spatial concentration of these new population groups. Thus, the main objectives of this work are to evaluate the degree and trend of residential segregation, as well as to locate and monitor the formation of concentration areas. To address them we compute two different but complementary segregation indexes (Dissimilarity and Theil) between 2003 and 2013 in the first case, and the local indicator of spatial association (LISA) Moran’s I in the second. The empirical strength of this paper is given by the available Spanish population data, which allows us to follow the trends on an annual basis, during a period of significant demographic and economic changes.

2. LITERATURE REVIEW AND BACKGROUND 2.1. Segregation and concentration of immigrant populations The spatial differentiation of population groups is one of the key characteristics of western cities (Hirschman, 1983). The reasons for such process may stem from the free will or preferences as well as from an imposition of the dominant group (Duncan and Duncan, 1955; Taeuber and Taeuber, 1965; Farley, 1977; Peach, 1996). In the case of the immigrant population in particular, reasons for segregation can by grouped into three categories (Kempen and Ozüekren; 1998): 1) Economic reasons linked to housing 1 This paper is result of the R+D Project “Diversidad, Segregación y Vulnerabilidad, Análisis sociodemográfico” (CSO2014-54059-R), diricted by Dr. Andreu Domingo, and financed by the Spanish Ministry of Economia y competitividad.

3

prices, the position of immigrants within the labour market and the location working opportunities; 2) Demographic reasons associated with processes of suburbanization and aging of autochthonous populations; 3) and those related to some particular “ethnic motivations”, as could be the development of informal networks of mutual help between members of the group.

One of the salient features of the process of spatial differentiation is the concentration of groups according to their social class and/or its place of birth. Concentration increases visibility and, under certain circumstances, it can even drive to the stigmatization of an area, hampering down the integration of newcomers (Alba et al., 1999). The so called neighbourhood effect (van Ham et al., 2012) also refers the impact of concentration over expectations, opportunities and living conditions of its residents. It is clear that when concentration results from an imposition, it generally goes along with other unfavourable circumstances for the well-being of those who suffer it. It has been studied in its relation with the geographical concentration of poverty (Massey, Gross and Shibuya, 1994); in its pervasive consequences over healht inequalities (Williams and Collins, 2001) and weight status (Chang, 2006); or in its inextricable relatedness with school segregation (Denton, 1996), to name a few examples. 2.2. The particular case of the Metropolitan Area of Barcelona (MBA) The MAB is a cosmopolitan place, and it already was at the beginning of the century. In 2003, people born in 176 different countries were living together within this high-density urban space, with 8 groups with populations over the 10 thousand, but none (except for the Spanish-born, obviously) over passing the 50 thousand people. A decade later, the number of countries represented is almost the same (179, see figure 1), but then we find two groups, those born in Ecuador and Morocco, with populations over 50 thousand people, and up to 19 groups with populations over 10 thousand people. The rapid diffusion of immigrant across the metropolitan region led some scholars to refer a process of ethnification of the metropolis (Garcia-Almirall et al., 2008; Miret, 2009), while the settlement of them in the different municipalities has contributed to its growing visibility across the whole area. The degree of residential segregation in the city of Barcelona, its metropolitan area, and the whole metropolitan region has been addressed by several scholars during the last decade (Martori and Hoberg, 2004; Bayona, 2007; Martori and Apparicio, 2011; García-Almirall et al., 2008). The general picture drawn by this "corpus", is that of low and decreasing levels of segregation, with some expected heterogeneity between different foreign-born groups according to their population size, arrival calendar and rates of internal mobility. The described trends framed the case of the MAB within a broader context of residential segregation in Spain (Domínguez et al., 2010) and in cities of Southern Europe. As it has been stated (Malheiros, 2002, Arbaci, 2008, Martínez and Leal, 2008), low levels residential segregation may not be taken as a clear indicator social incorporation within these cities, mainly due to some historical features of the urban evolution, and development of welfare state in these countries.

4

FIGURE 1. Cartogram of Countries of birth of the immigrant population living in the Metropolitan Area of Barcelona, 2013

Source: Own elaboration, data from the Municipal Registers, 2013 (INE). Software: ScapeToad. ST uses the Gastner/Newman [2004] diffusion-based algorithm to adapt map surfaces to user-defined variables without altering their topological relations. Despite the low degree of residential segregation, the MAB has not been exempt of the formation of concentration areas, becoming a major concern when it comes to the design and implementation of integration policies (Zapata, 2001). Within the MAB, major concentrations of foreign-born population were found in the old town of the municipality of Barcelona, gateway neighbourhood for the international migration (Aramburu, 2001), and in its neighbour municipalities (Galeano et al., 2014; Sabater et al., 2013 or Martori and Apparicio, 2011) 3. OBJETIVES This study builds on the previous work in this area (Peach, 1997; Stillwell and Van Ham, 2010; Sabater et al., 2012) and it is considered to be an important contribution to the segregation debate because it considers the process of settlement and clustering of immigrant groups at small spatial areas (census tracts) over time and space. More specifically, the aims of this paper are:

1) Assess the degree of residential segregation between different foreign-born populations and natives; as its trend during the boom and bust of international migration into the Metropolitan Area of Barcelona (2003-2013).

2) Evaluate and monitor the formation and evolution of clustering areas of both, sub-representation and concentration, for different immigrant groups at the census tract level for the whole MBA.

4. DATA AND METHODOLOGY The Metropolitan Area of Barcelona is the territorial entity composed by the municipality of Barcelona plus other 35 surrounding municipalities2. Recognized us 2 A map of the composition by municipalities of the MBA can be found in the annex.

Forign-born population

MBA 2013

5

such since 2011 after the fusion of three pre-existent agencies (Union of Municipalities, the Transport Authority and the Environment Authority), it has an extension of 636 km2 and a population, in 2013, of 3,228,569 people (four out of ten people living in Catalonia, Spain). For our analysis we use data from the Municipal Register of Inhabitants (Padrón Continuo de Población) between 2003 and 2013. The Municipal Register is the administrative register where municipality inhabitants are recorded. The respective town councils are responsible for its formation, maintenance, revision and custody and its update is obtained from the revision of the municipal register referring to 1 January of each year, which is approved by the Government at the Instituto Nacional de Estadistica's (INE) request, after the favourable report by the Registration Board. This data set allows us to know annually the basic socio-demographic characteristics (sex, age, nationality and place of birth) of the population of each municipality at the census tract level. In methodological terms, we employ two different approaches that allow us to get, in a first step, the general picture of the spatial incorporation of foreign-born populations at the regional level (MAB), and secondly, the one related with the concentration of those groups at census tracts level. One difficulty encountered by our analysis is the variability of the number of census tracts that make up the MAB between 2003 and 2013. Its number is always variable, since their creation or redefinition is governed by electoral criteria (theoretically, they have to gather between 500 and 2,000 voters), and changes depending on the evolution of the population. In a context of population growth, the census tracts used in this work increased from 2,531 in 2003 to 2,576 until 2009. In 2010 there was a noticeable decline in census units mainly due to redefinition of those of the municipality of Barcelona (with a strong reduction from 1,482 to 1,061), according to the order of the Neighbourhood Act passed in 2004 and implemented in 2009, by 2013 the MAB is composed by 2,157 census tracts. Residential segregation is a multidimensional phenomenon (Massey and Denton, 1998); in this work we focus mainly in the unevenness of the spatial distribution for different populations, grouped by their region of birth (Latin-America, Western Europe3, Eastern Europe4, Africa and Asia), with respect to that of the Spanish-born. In order to assess the degree and evolution of residential segregation, we compute the index of dissimilarity (ID) for the whole of MAB, for the capital municipality of Barcelona and for the rest of the MAB separately. The index of dissimilarity (D), formulated by Duncan and Duncan (1955) was calculated using the following formula:

𝐷 =12��

𝑥𝑖𝑋−𝑦𝑖𝑌�

𝑛

𝑖=1

∗ 100

3 According to the classification of the Statistic Division from the United Nations (2011), Western Europe includes: Germany, Andorra, Austria, Belgium, Denmark, Finland, France, Ireland, Island, Italy, Liechtenstein, Malta, Monaco, Norway, Netherland, Portugal United Kingdom, San Marino, Sweden, Switzerland and the Vatican city. 4 According to the classification of the Statistic Division from the United Nations (2011), Eastern Europe includes: Albania, Bulgaria, Chipre, Hungry, Poland, Romania, Ukraine, Latvia, Moldova, Belarus, Georgia, Estonia, Lithuania, Czech Republic, Slovak Republic, Bosnia and Herzegovina, Croatia, Slovenia, Armenia, Russia, Serbia and Montenegro and Macedonia

6

Where 𝑥𝑖 is the population of a 𝑋 type within the 𝑖 area, e.g. census tracts, municipalities; 𝑦𝑖 is the population of a 𝑌 type within the 𝑖 area, 𝑋 is the total 𝑋 population of the large geographic entity for which the index is being calculated, 𝑌 is the total 𝑌 population of the large geographic entity for which the index is being calculated. We multiplied the resulting value of D by 100 to facilitate the interpretation of the results as the percentage of the group under consideration that would need to change their place of residence to replicate the spatial distribution of the reference group (the Spanish-born population in this case). The resulting value of the index allows us to know, in a synthetic way, the degree of spatial assimilation of immigrant populations with respect to the reference group, as well as the trend over time, but tells us nothing regarding the specific location where segregation is taking place. In order to disentangle if the observed level of segregation for the MAB results from that taking place within or between municipalities, we implement the Theil Index (H) which, because of its properties, can be additively decomposed into contributions from different geographical levels (Lichter, D. et al. 2015; Farrell, 2014). H is based on the entropy concept of diversity (Theil, 1972), which if we take two population groups at the time (e.g., Latin-Americans and Spaniards) can be formulated as follows:

𝐸 = 𝑝𝑖 ln1𝑝𝑖

+ (1 − 𝑝𝑖) ln1

1 − 𝑝𝑖

Where 𝑝𝑖 is the proportion of Latin-Americans. Computed over two population groups E reaches a maximum value of 0.69 when both groups are equally represented and a minimum value of 0 if there is only one group in the area under analysis. Taking the Metropolitan Area of Barcelona as the total geography (U) and census tracts as subunits (B) the Theil Index (H) can be expressed as:

𝐻𝐵 𝑈 =1

𝑁𝑈𝐸𝑈 �𝑁𝑏(𝐸𝑈 − 𝐸𝑏)𝐵

𝑏=1

Where 𝑁𝑈 is the total population of the MAB, and 𝑁𝑏 is the total population of the census tract. 𝐸𝑈 is the overall entropy of the MAB and 𝐸𝑏 is the entropy of the census tract. To disentangle which part of the observed segregation for the MBA in taking place within municipalities and which between municipalities we decompose 𝐻𝐵 𝑈 as follows:

𝐻𝐵 𝑀 =1

𝑁𝑈𝐸𝑈�𝑁𝑏(𝐸𝑀 − 𝐸𝑏)

𝐵

𝑏=1

to get the within municipalities part of the overall segregation and

𝐻𝑀 𝑈 =1

𝑁𝑈𝐸𝑈�𝑁𝑚(𝐸𝑈 − 𝐸𝑚)

𝑀

𝑏=1

7

for the part of the overall segregation that comes from the one between municipalities. In second term, we compute the local indicator of spatial association Moran’s 𝐼 (Anselin, 1995), to isolate clustering areas of both concentration and/or sub-representation of different foreign-born populations within a map. The formula for 𝐼𝑖 is:

𝐼𝑖 = 𝑧𝑖�𝑤𝑖𝑗𝑧𝑗𝑗

Where 𝑧𝑖 is: 𝑧𝑖 = 𝑥𝑖 − 𝑋�

where 𝑧𝑖 is the original variable in its “deviation form”; 𝑥𝑖 is the proportion of the population of the type 𝑥 living in the census tract 𝑖 and 𝑋� is the mean proportion of the population of all areas in ethnic group 𝑥; 𝑤𝑖𝑗 is the spatial proximity weight for areas 𝑖 and 𝑗 (defined as a the union of a proximity matrix of 500 metres with a supplementary k-nearest neighbour weigh matrix where K=1) and 𝐼𝑖 is the value (distributed as 𝑍) for area 𝑖 . As 𝐼 has the same distribution as 𝑍 , statistically significant clusters of both concentration and relative absence can be identified directly by mapping its values. In the cartography presented here we established a threshold of statistical significance over 95% (values above 1.96 or below -1.96) for cluster of concentration as well as for those of sub-representation. 5. RESULTS: 5.1. Residential Segregation, 2003-2013 Residential segregation can be defined as the degree to which various population groups share, or not, a common residential space. It's worth noting also the double nature of this phenomenon: On the one hand, it results from a particular setup of a given society; but on the other, it could be also a mechanism by which inequalities, of any kind, can get perpetuated or even amplified (Maloutas and Fujita, 2012). One of the most extended methods used by researchers to quantify this phenomenon has been the index of dissimilarity which compares, at a given time, the spatial distribution of a group defined on the basis of some shared characteristic (the region of birth in our case) to that of other group taken as reference (here, the Spanish-born population). The resulting value of the index allows us to know, in a synthetic way, the degree of spatial assimilation of immigrant populations with respect to the reference group, as well as the trend over time. Figure 2 depicts the evolution of the dissimilarity index for five population groups between 2003 and 2013. Residential segregation, as measured by this index, ranges between moderate-to-low values for all groups (below the 50 points in all cases in 2014). The evolution also shows an increasing spatial assimilation, except for that of the population born in countries of Western Europe (mainly Italians, French and Germans), in particular those living in the municipality of Barcelona. Decreasing residential segregation over the period 2003-2013 reflects both the demographic increase and spatial diffusion of groups under analysis and, consequently, the increase of shared residential areas with the Spanish-born population. This result is consequence of the general balance between the degree of concentration and under-representation of

8

immigrant groups in the MAB in relation to the spatial distribution of the population born in Spain. FIGURE 2. Dissimilarity index for different immigrant population, 2003-2013

Source: Own elaboration with data from the Municipal Registers, 2003-2013 (INE) The general trend, as well as the degree, is also reflected by the Theil Index (Figure 3 and 4), with Latin-Americans and Eastern Europeans achieving the lower values of residential segregation in 2013. In the case of the later, the sharp decrease of the observed values is mainly due to the huge increase in the number of this population as a result of intense migratory flows unchained by the inclusion of Romania and Bulgaria in the EU in 2004. The index also reflects the increasing residential segregation of Western-Europeans as well as the higher degree for Africans and, even more, Asians. The Theil index give us complementary information about the administrative level in which the observed metropolitan residential segregation for each group is taking place; it tell us the degree in which segregation is an outcome of the situation within or between municipalities. (Table 1). In the case of both Latin-Americans and Western European 60% of the observed segregation stems from the one that takes place within municipalities, and the share remains almost constant during the whole period of analysis. In the case of Eastern-Europeans, Africans and Asians the share is around 78% at the end of period, and the decrease during the decade reflects, in a way, the diffusion of these populations among municipalities of the MAB and their higher levels of intra-municipal concentration and separation from the Spanish-born population. The largest decrease is among Africans (from 90% to 79%), with Asians showing a milder decrease (82% to 77%) and Eastern Europeans on an intermediate situation (85% to78%). The general decline in the weight that intra-municipal segregation plays over the metropolitan one is the result of diverse trends: for Latin-Americans, it is an outcome of the decrease in both components of metropolitan segregation; as is the case also for Eastern European, but in a sharper way. Instead, for the African and Asian populations, it results from a mixed process: the decrease on intra-municipal segregation, but the increase of the inter-municipal one. Western Europeans show a steady increase in both components of metropolitan segregation

9

FIGURE 3. Theil index for different immigrant population groups, 2003-2013

Source: Own elaboration with data from the Municipal Registers (INE), 2003-2013.

FIGURE 4. Descomposición de la evolución de la segregación en la MAB (2003-2013)

Source: Own elaboration with data from the Municipal Registers (INE), 2003-2013.

10

TABLE 1. Share of the observed metropolitan segregation occurring within municipalities

Source: Own elaboration with data from the Municipal Registers (INE), 2003-2013.

5.2. Spatial Clustering of Immigrant Populations5 In order to assess changes occurred in the spatial configuration of the metropolitan space, in relation with the composition of its population, we proceed to the location of both clusters of concentration and sub-representation of different immigrant groups at the beginning (2003) and the end of our period of analysis (2013). A concentration cluster can be defined as a place where neighbours share some characteristic, in our case the region of birth of their inhabitants, and this feature comes to be over-represented relative to its distribution within the rest of the territory of reference. On the contrary, a cluster of sub-representation is composed by those neighbour areas where the same characteristic is underrepresented, relative to its distribution within the rest of the territory of reference. The local indicator of spatial association Moran (I) allows us to locate those clusters by mapping the resulting values. A central component of both this indicator as of any general definition of a cluster is the neighbourly relation we establish between spatial units, defined at 500 meters in this case with the complement of nearest neighbour for those census tract with no connection from the previous definition. It should also be noted that the final value of each spatial unit (here census tracts) in this type of analysis is the result between the ratio of the value of the unit concerned and the values of its neighbours. Figure 4 to Figure 8 present the cartography resulting from the exercise of localization of clusters of concentration and sub-representation for the years 2003 and 2013 for each immigrant group. First, as expected due to the volume and intensity of migration flow, we corroborate, for all immigrant groups an expansion and to some extent relocation, of the clustering areas of concentration of each group. We start by highlighting what can be generalized to all population groups: for both years, the neighbourhoods Raval and Gòtic (located in the inner city of Barcelona) are configured as part of the city where there are located cluster of concentration of all groups of foreigners without exception.

5 Find maps about the neighbourhoods of the municipality of Barcelona in the ANNEX

Intra Between Total % intra Intra Between Total % intraLatin-America 0,08 0,05 0,13 61,2 Latin-America 0,07 0,05 0,12 60,8Western-Europe 0,06 0,04 0,10 61,0 Western-Europe 0,09 0,06 0,16 59,2Eastern-Europe 0,17 0,03 0,20 86,0 Eastern-Europe 0,06 0,02 0,08 78,7Africa 0,17 0,02 0,18 89,8 Africa 0,13 0,04 0,17 78,8Asia 0,29 0,06 0,36 82,4 Asia 0,23 0,07 0,30 77,0

Intra Between Total % intra Intra Between Total % intraLatin-America 0,00 0,00 -0,01 -0,43 Latin-America -4,88 -3,16 -4,22 -0,70Western-Europe 0,03 0,03 0,06 -1,82 Western-Europe 52,62 64,63 57,30 -2,98Eastern-Europe -0,11 -0,01 -0,13 -7,30 Eastern-Europe -65,45 -42,56 -62,24 -8,49Africa -0,03 0,02 -0,02 -10,97 Africa -20,85 86,94 -9,83 -12,22Asia -0,06 0,01 -0,05 -5,39 Asia -20,66 10,86 -15,11 -6,54

2003 2013

2003-2013 (absolute) 2003-2013 (relative)

11

Secondly, by 2013, clusters of concentration of different groups are spread in 20 out of the 36 municipalities of the MAB. In this regard, the largest population diversity and centrality of the municipality of Barcelona is reflected in being the only municipality in which we have found concentration areas of all population groups. After Barcelona, the coastal city of Castelldefels gathers concentration areas of Latin Americans, Western Europeans, Eastern Europeans and Africans, a similar situation founded in the city of Badalona, in this case with concentrations of Asians, Africans and Eastern Europeans. Latin-Americans, represented in Figure 5, are the largest immigrant group in the MAB, accounting for half of the total foreign-born population in both years (2003 and 2013), but having doubled its population from 164 to 320 thousand people. Aside from the above mentioned clusters in the Raval and Gòtic, the other great cluster of concentration of this population in 2003 was located in some neighbourhoods of the central city, like Fort Pienc. By 2013 we can trace its extension following the Meridiana Avenue, reaching part of the adjacent neighbourhoods (El Clot, Camp de l’Arpa, Navas, la Sagrera, Congrés, Sant Andreu, Bon Pastor and Ciutat Meridiana), all of them with a household income below the average. This exercise also allows us to observe the expansion, by 2013, of the Latin-American cluster in the town of Castelldefels, the persistence of that located in the high-income city of Sant Cugat del Vallès and, conversely, the disappearance of cluster areas that appeared in 2003 in some peripheral municipalities characterised by a dispersal habitat like Cervelló. We should also note that some of the Latin-American clusters are those with the highest degree of residential exclusivity, since often the members of this group represent more than 65% of foreigners living in those areas (67% in Sant Cugat del Vallès and L'Hospitalet de Llobregat, or up to 70% in the case of Esplugues de Llobregat) (Galeano and Bayona, 2015). FIGURE 5. Moran’s I for Latin-American immigrant populations, Metropolitan Area of Barcelona, 2003-2013

2003 2013

Source: Own elaboration with data from the Municipal Registers (INE), 2003 and 2013 The second largest immigrant group in the MAB it's that of the Asians (figure 6). This population group is even more heterogeneous than the one of Latin-Americans, mixing cultures and geographies as diverse as those of people born in China, Pakistan, Philippines, Bangladesh and India among many others. The clusters of concentration of this group -which on average, between 2003 and 2013 represents 14% of total foreign population, and has multiplied by 2.61 its population: from 39,585 to 103,476 people- are found, in addition to the aforementioned quarters of Raval and the Gòtic (where the population is mostly born in Pakistan, the Philippines and Bangladesh) at the

12

municipalities of Badalona (neighbourhoods of El Remei, Sant Roc, Artigas, La Pau, Nova Lloreda and Puigfred) and Santa Coloma de Gramenet (neighbourhood of El Fondo), where the Chinese and Pakistani are the predominant foreign-born groups. The case of Asians illustrates the situation in which cluster areas get consolidated and expanded after intense population growth. In this case, ancient concentration patterns are newly reproduced in the first metropolitan ring. FIGURE 6. Moran’s I for Asians immigrant populations, Metropolitan Area of Barcelona, 2003-2013.

Source: Own elaboration with data from the Municipal Registers (INE), 2003 and 2013. In the case of the population born in countries of Western Europe (figure 7), population growth (from 45,859 to 81,351 people between 2003 and 2013) has been accompanied by a clear territorial reconfiguration of clustering areas, always located within the wealthier areas of the metropolis. While in 2003 we found them at the outlying municipalities of Castelldefels, Gavà, Cervelló, Sant Cugat del Vallès and Sant Just Desvern, plus on the wealthy quarters of the city of Barcelona (Sant Gervasi, Les Tres Torres, Sarrià and Pedralbes) and at the neighborhoods of Gòtic and Raval, in 2013 we see the great cluster of this population is located in the heart of the city of Barcelona (Barri Gòtic, Raval, Eixample, Gracia and Sant Gervasi). By 2013 the clustering areas of this population have also extend across other neighbourhoods of the city (Barceloneta, Vila Olímpica, Poble Nou, Poble Sec and Diagonal Mar). At the rest of municipalities of the MAB we've just found another cluster located in Castelldefels and Gavà and another between Esplugues de Llobregat and Sant Just Desvern, where it can be argue that was the location of the German and American schools, and the proliferation of residential complexes in the latter that encouraged the settlement of this group within these areas. The differential pattern of settlement of Western European is reflected in the fact that L'Hospitalet de Llobregat, Santa Coloma de Gramenet, Badalona and the less wealthy areas of the north-east of the municipality of Barcelona appear like cluster areas of significant sub-representation for this group.

13

FIGURE 7. Moran’s I for Western-Europeans immigrant populations, Metropolitan Area of Barcelona, 2003-2013.

Source: Own elaboration with data from the Municipal Registers (INE), 2003 and 2013. The African-born population represented 15.4% (49,421 people) of the immigrant population living in the MAB in 2003 (figure 8). Ten year later its population grew up to 69,854, but its share among the total foreign-born decreased to 11.3%, due to the massive increase in population of all other groups. We find clustering areas of Africans in 15 different municipalities, which reflects the high intra-municipal concentration of this group. The picture between 2003 and 2013 is quite similar, although characterized by an extension of the areas of concentration towards the low-income municipalities in the north-east Besòs area. Africans are poorly represented in the central city but, as other immigrant groups, concentrated at the Gotic and Raval quarters. In 2013 Trinidad Vella also appears as a clustering area, which it's not surprising due the long-time presence of the Moroccan community in that area. At the rest of the MAB, concentrations are located in most central cores of the municipalities of the southern Llobregat (in some cases with significant presence of Moroccans since the early nineties, related with agricultural industry) and in other towns close to Barcelona as Santa Coloma de Gramenet, Badalona or Montcada i Reixac (all of them in the Besos area). By contrast, much of the city of Barcelona appears as being an area of sub-representation for this group. FIGURE 8. Moran’s I for Africans immigrant populations, Metropolitan Area of Barcelona, 2003-2013.

Source: Own elaboration with data from the Municipal Registers (INE), 2003 and 2013.

14

Lastly, Eastern European population (figure 9) grew from 16,902 to 52,474 people between 2003 and 2013, rising it share among the foreign-born from 5.3% to 8.5%. It’s important to bear in mind the different socio-economic profile of members of this group composed mainly by Romanians and Bulgarians -with a clear labour migrant profile- but also by large population born in Russia -with a residential behaviour which can be assimilated to that of the Western European population-. Population cluster of this group are found (aside from those in the Raval and Gòtic) in the neighbourhood of the coastal district of Sant Martí (Llacuna, Vila Olimpica, Poble Nou, Diagonal Mar and Besòs) and Ciutat Vella (Barceloneta and Parc). Outside Barcelona concentration areas are few, located as other groups in the coastal municipalities of Castelldefels and Gavà and also in some low-income municipalities of Ripollet and Cerdanyola del Vallès. FIGURE 9. Moran’s I for Eastern-Europeans immigrant populations, Metropolitan Area of Barcelona, 2003-2013.

Source: Own elaboration with data from the Municipal Registers (INE), 2003 and 2013.

6. CONCLUSIONS The main objective of this paper was to assess the process of territorial settlement of recent immigrant groups in the Metropolitan Area of Barcelona, taking into the account the degree of spatial assimilation of each group as well as evaluating the formation of statistical significant clustering areas. The results obtained in this work allows us to state that, in general and despite the large increase of the foreign-born population living in the MAB, the levels of residential segregation of all immigrant groups, with respect to the Spanish-born population, are moderate-to-low, and have tended to decrease between 2003 and 2013 (with the only exception of that of population born in Western Europeans countries). Although we have not presented empirical evidence that allow us to disentangle the causes for such process, we can convincingly argue that there are three major issues taking part in the reduction of residential segregation within the MAB: 1) In first place the aforementioned increases in population of all groups and its impact on residential segregation values (Simpson, 2004); on second and third place, a linked process that involves the 2) intense suburbanization of the Spanish-born population (especially between 2000 and 2007) (Pujadas et al, 2012) and 3) the high internal mobility of all immigrants groups (Bayona and Gil, 2012). All these interacting phenomenon result in the widespread territorial diffusion of the new groups, and in the sharp increase on mixed residential areas.

15

We complemented the information given by the dissimilarity index with the one obtained by computing the Theil Index, by which we can state that the major part of the observed segregation is due to the unevenness of the spatial distribution within municipalities, notoriously in the case of Eastern Europeans, Africans and mainly Asians. In the case of Africans and Asians the general decrease in their degree of metropolitan segregation involves a mixed trend, with a decline of the level of segregation observed within municipalities and an increase of the one between municipalities. The interplay between the two trends gives us an idea of how segregation processes are increasingly taking a metropolitan dimension. Finally, and in order to confront the lack of spatial dimension of both indexes, we proceeded to locate clustering areas of each population group. Mapping the values of Local Moran’s revealed as a very effective way to monitor the formation and evolution of clustering areas of both concentration and sub-representation at the census tract level, despite, and taking into account, the huge population growth experienced by all foreign-born groups during our period of analysis. In general terms what we found is the consolidation in 2013 of the concentration areas that were already there in 2003, their expansion (markedly in the case of the Latin-American population), as well as some sort of reconfiguration (as in the case of clustering areas of the Western-European population).

16

7. BIBLIOGRAPHY ALBA, R.D., LOGAN, J.R., STULTS, B.J., MARZAN, G. and ZHANG, W. (1999), «Immigrant Groups in the Suburbs: A Re-examination of Suburbanization and Spatial Assimilation», American Sociological Review, 64, pp. 446-60.

ANSELIN L. (1995), «Local indicators of spatial association – LISA», Geographical Analysis, 27, pp. 93–115.

ARAMBURU, M. (2001), «El mito de la “huida” autóctona. El caso de Ciutat Vella, Barcelona» Scripta Nova, Revista Electrónica de Geografía y Ciencias Sociales, Núm. 94 (63).

ARBACI, S. (2008), «(Re)viewing ethnic residential segregation in Southern European cities: housing and urban regimes as mechanisms of marginalization» Housing Studies, 23, pp. 589–613.

BAYONA J. and GIL-ALONSO F. (2012), «Suburbanisation and International Immigration: the case of the Barcelona metropolitan region (1998-2009)», Tijdschrift voor Economische en Sociale Geografie 103(3), pp. 312-329.

BAYONA, J. (2007), «La segregación residencial de la población extranjera en Barcelona: ¿una segregación fragmentada?» Scripta Nova. Revista Electrónica de Geografía y Ciencias Sociales. Vol. XI, (235).

BAYONA, J. and LÓPEZ-GAY, A. (2011), «Concentración, segregación y movilidad residencial de los extranjeros en Barcelona», Documents d’Anàlisi Geogràfica, 57 (3): pp. 381-412.

CHACÓN-RODRÍGUEZ, L., (2002), «La Formación de la España Inmigrante: Mercado y Ciudadanía», Reis, 97(2), pp. 95-126.

CHANG, V., (2006), «Racial Residential Segregation and Weight Status among US Adults», Social Science and Medicine, Vol. 63, pp. 1289-1303.

DENTON, N., (1996), «The Persistence of Segregation: Links Between Residential Segregation and School Segregation». Minnesota Law Review., 80, pp. 795-824..

DOMÍNGUEZ, J.; PARREÑO, M. and DÍAZ, R. (2010), «Inmigración y ciudad en España: integración versus segregación socio-territoriales» Scripta Nova. Revista Electrónica de Geografía y Ciencias Sociales, Vol. XIV, 331 (50).

DUNCAN, D. & DUNCAN, B., (1955), «A Methodological Analysis of Segregation Indexes», American Sociological Review, Vol. 20, pp. 210-217.

FARLEY, R., (1977), «Residential Segregation in Urbanized Areas of the United States in 1970: An Analysis of Social Class and Racial Differences», Demography, Vol. 14, pp. 497-518.

FARRELL, C.R. (2014), «Immigrant suburbanisation and the shifting geographic structure of metropolitan segregation in the United States», Urban Studies, doi:10.1177/0042098014558537.

FREY, W.H. (2001), Melting Pot Suburbs: A Census 2000 Study of Suburban Diversity. Center on Urban and Metropolitan Policy, Brookings Institution, Washington, DC.

GALEANO, J. and BAYONA, J. (2015), «Assentament territorial de la població estrangera a l’Àrea Metropolitana de Barcelona en el segle XXI», In: DOMINGO, A. (Coord.) Migracions del segle XX i XXI a Catalunya. Una mirada Candeliana.

17

Col·lecció Ciutadania i immigració, n. 11, Departament de Benestar Social i Família, Generalitat de Catalunya, pp. 95-121.

GASTNER, M.T. and NEWMAN, M. (2004). «Diffusion-based method for producing density equalizing maps». Proceedings of the National Academy of Sciences of the United States of America, 101(20), pp. 7499-7504.

GARCÍA-ALMIRALL, P., FULLAONDO, A. and FRIZZERA, A., (2008), «Inmigración y Espacio Socio-Residencial en la Región Metropolitana de Barcelona», Ciudad y Territorio, Estudios Territoriales, Vol. 40, pp. 727-741.

HIRSCHMAN, C., 1983, «America's Melting Pot Reconsidered», Annual Review of Sociology, Vol. 9, pp. 397-423.

HAM, M. Van, MANLEY, D., BAILEY, N., SIMPSON, L. and MACLENNAN, D. (eds.) (2012). Neighbourhood Effects Research: New Perspectives. London: Springer

KEMPEN, R. van, and ÖZÜEKREN, S. (1998), «Ethnic Segregation in Cities: New Forms and Explanations in a Dynamic World» Urban Studies, 35, (10), pp. 1.631-1.656. LICHTER, D., PARISI, D. and TAQUINO, M. (2015), «Toward a New Macro-Segregation? Decomposing Segregation within and between Metropolitan Cities and Suburbs», American Sociological Review, vol. 80(4), pp. 843-873. LOGAN, J.R., STULTS, B. and FARLEY, R. (2004), «Segregation of minorities in the metropolis: two decades of change», v. 41(1), pp. 1-22.

LOPEZ-GAY, A. (2014), «Population growth and re-urbanization in Spanish inner cities: The role of internal migration and residential mobility», Revue Quetelet/Quetelet Journal, v. 2 (1), pp. 67-92.

MALHEIROS, J. (2002), «Ethni-cities: residential patterns in the Northern European y Mediterranean metropolises–implications for policy design», International Journal of Population Geography, 8,pp. 107–134.

MALOUTAS, T. and FUJITA, K. (Eds.) (2012), Residential Segregation in Comparative Perspective. Making Sense of Contextual Diversity, City and Society Series, Ashgate, Farnham, Reino Unido.

MARTÍNEZ, A. and LEAL, J (2008) «La segregación residencial, un indicador espacial confuso en la representación de la problemática residencial de los inmigrantes económicos: el caso de la Comunidad de Madrid» ACE, 3, pp. 53-64.

MARTORI, J.C. and APPARICIO, P. (2011) «Changes in Spatial Patterns of the immigrant population of a Southern European Metropolis: the case of the Barcelona Metropolitan Area (2001-2008)». Tijdschrift voor Economische en Sociale Geografie, 102 (5), pp. 562-581.

MARTORI, J.C. and HOBERG, K. (2004) «Indicadores cuantitativos de segregación residencial. El caso de la población inmigrante en Barcelona». Scripta Nova. Revista electrónica de geografía y ciencias sociales, vol. VIII, n. 169.

MASSEY, D. and DENTON, N., (1988), «The Dimensions of Residential Segregation», Social Forces, Vol. 67, pp. 281-315.

MASSEY, D., GROSS, A. & SHIBUYA, K. (1994), «Migration, Segregation, and the Geographic Concentration of Poverty», American Sociological Review, Vol. 59, pp. 425-445.

18

MIRET, N., (2009), «Inmigración y metropolización en Barcelona», AREAS. Revista internacional de Ciencias Sociales, Vol. 28, pp. 103-116.

PEACH, C., (1996), «The meaning of Segregation», Planning Practice and Research, Vol. 11, pp. 137-150.

PEACH, C., (2001), «The Ghetto and the ethnic enclave», International Seminar on Segregation in the City. Lincon Institute, pp. 1-21.

PUJADAS-RÚBIES I. (2009), «Movilidad residencial y expansión urbana en la región metropolitana de Barcelona, 1982-2005», Scripta Nova: Revista Electrónica de Geografía y Ciencias Sociales 13, (290).

SABATER, A.; BAYONA, J. and DOMINGO, A. (2012), «Internal migration and Residential patterns across Spain after unprecedented international migration». In: FINNEY, N.; CATNEY, G. (ed.). Minority Internal Migration in Europe, Surrey: Ashgate Publishing, International Population Studies Series, cap. XIV, pp. 293-311. ISBN 978-1-4094-3188-6.

SIMPSON, L. (2004), «Statistics of Racial Segregation: Measures, Evidence and Policy». Urban Studies, Vol. 41, pp. 661-681.

STILLWELL, J. and VAN HAM, M. (2010), Ethnicity and Integration. Understanding Population Trends and Processes. London/New York: Springer.

TAEUBER, K. and TAEUBER, A. (1965), Negroes in Cities: Residential Segregation and Neighborhood Change. Chicago: Aldine Publishing Company.

THEIL, H. (1972), Statistical Decomposition Analysis. North-Holland Publishing Co.

WILLIAMS, D. and COLLINS, C. (2001), «Racial Residential Segregation: A Fundamental Cause of Racial Disparities in Health», Public Health Reports, Vol. 116, pp. 404-416.

ZAPATA, R. (2001), «La relació entre els immigrants i les Administracions: Onze temes bàsics per a debatre polítiques d’integració», Papers de la Fundació, 122

19

ANNEX Figure 1. METROPOLITAN AREA OF BARCELONA: MUNICIPALITIES

Source: Own elaboration

Figure 2. MUNICIPALITY OF BARCELONA: NEIGHBOURHOODS

Source: Own elaboration

1. Raval (a) 29. Putxet i Farró2. Raval (b) 30. La Salut-Vallcarca-El Coll3. Gòtic 31. Vila de Gràcia (a)4. Barceloneta 32. Vila de Gràcia (b)5. St. Pere, Sta. Caterina i Ribera 33. Camp d'en Grassot i Gràcia

Nova6. Fort-Pienc 34. Baix Guinardó i Can Baró7. Sagrada Família (a) 35. Guinardó8. Sagrada Família (b) 36. Horta-Font d'en Fargues9. Dreta Eixample (a) 37. Carmel10. Dreta Eixample (b) 38. Vall d'Hebron, Montbau, St.

Genís i Teixonera11. Antiga Esquerra Eixample 39. Vilapicina12. Nova Esquerra Eixample (a) 40. Porta13. Nova Esquerra Eixample (b) 41. Turó de la Peira, Guineueta i

Can Peguera14. Sant Antoni 42. Canyelles-Roquetes15. Poble_Sec 43. Verdum-Prosperitat16. Font de la Guatlla (a) 44. Trinitat Nova, Torre Baró,

Ciutat Meridiana, Vallbona17. Font de la Guatlla (b) 45. Trinitat Vella, Baró de Viver i

Bon Pastor18. La Bordeta i Hostafrancs 46. Sant Andreu (a)19. Sants-Badal 47. Sant Andreu (b)20. Sants (a) 48. La Sagrera21. Sants (b) 49. Congrés-Navas22. Les Corts (a) 50. Camp de l'Arpa23. Les Corts (b) 51. El Clot24. Pedralbes, Maternitat i Sant Ramón 52. Parc, Llacuna i Vila Olímpica25. Sarrià-Vallvidrera 53. Poblenou26. St. Gervasi Bonanova-Tres Torres 54. Diagonal Mar i Front Marítim27. St. Gervasi-Galvany (a) 55. Besòs i Maresme28. St. Gervasi-Galvany (b) 56. Sant Martí de Provençals

57. Verneda i La Pau