Foreign Migration, Urban Growth and Suburbanisation Dynamics in Large Spanish Metropolitan Areas

Transcription

1 Enhr Conference July, Toulouse Foreign Migration, Urban Growth and Suburbanisation Dynamics in Large Spanish Metropolitan Areas Jordi Bayona-i-Carrasco Human Geography Department, Universitat de Barcelona Fernando Gil-Alonso Human Geography Department, Universitat de Barcelona Abstract This paper analyses the impact of international migration on the development and population composition of the Spanish largest urban areas, focusing on foreigners participation in suburbanization dynamics. In this last decade, Spain has had the largest international migration inflows in Europe. The proportion of foreign residents has therefore increased from a trifling 2.3% in 2000 to today s 12.2%. According to 2010 data, the percentage of foreigners living in the central cities of the approximately fifteen large metropolitan areas with more than half a million inhabitants, ranges from 17.5% in Barcelona to 1.7% in Cadiz. The paper seeks: 1) to provide an overview of recent population changes in Spanish metropolitan areas; 2) to distinguish stages in them; 3) to search for urban development timing differences, as well as variations of their respective central city s role and of the type of fringe areas they have; and finally, 4) to analyse residential mobility dynamics. Keywords: Foreign immigration, suburbanisation, urban growth, metropolitan areas, Spain. Introduction Spain s population has undergone an abrupt increase due to an intense, brief and unexpected foreign immigration rise. In a decade, the country reached, from less that 40 million inhabitants, -considered in some early 1990 s population projections (Instituto de Demografía, 1994) as Spain s ceiling-, to 47 million residents. These 6 extra million inhabitants represent the highest absolute and relative (more than 14%) growth since the Additionally, it should also be taken into account that this increase coexisted with extremely low fertility and mortality rates. Even though such a growth has an impact over the whole territory, certain areas have been particularly affected due to foreigner s uneven spatial distribution. The paper focuses on its effects on the main Spanish metropolitan areas, which are districts recently undergoing major changes (Feria and Albertos, 2010). Though, as we will see, many have been considered by foreign immigrants as attraction points (but also only as entry doors, therefore settling elsewhere), other cities have hardly received any demographic impact. In the mid 1990 s, the Spanish urban system was at a maturity stage, and suburbanisation and metropolitan

2 Mixité : an urban and housing issue? 2 expansion processes dominated. In other words, most metropolitan centres were losing population, their residents were progressively ageing and, urban peripheries were rapidly expanding. Then, particularly from 2000 onwards, this development was abruptly interrupted by the massive arrival of foreign migrants, beginning a new urban demographic growth cycle, which present authors have studied for the four major Spanish metropolitan areas: Madrid, Barcelona, Valencia and Seville (Bayona and Gil-Alonso, 2008; Bayona et al., forthcoming). Population decline in their central cities started to diminish and some are even presently once again gaining population as foreigners are compensating the constant local origin population loss. Meanwhile suburbanization processes and flows, mainly made up of Spanish people though increasingly also of immigrants, grew, accelerating the phenomenon. This demographic and geographic expansive cycle ends when the global economic crisis -worsened by a local real estate market collapse after the inflationist period- strikes Spain. Even though foreigner inflows have, since then, decreased in Spain as a whole, their impact on the largest urban areas is probably very heterogeneous. As this latter period has already lasted for some years, we belief that some conclusions can start to extracted both from changes introduced by high immigration years and from the effects of the economic crisis on migration flows. Therefore, we aim to: 1) to provide an overview of recent population changes in Spanish metropolitan areas; 2) to distinguish stages in their development (urban core population decrease, new growth period due to international immigration, new suburbanization stage by foreigners, and the impact of the economic crisis); 3) to search for urban development timing differences between metropolitan areas, among their respective central city s role and in the type of fringe areas; and finally, 4) to analyse residential mobility dynamics, taking differences between foreigners and Spaniards choice of place of residence into account. Theoretical framework The classical cyclical urbanisation model built by Van den Berg et al. (1982) is widely accepted by urban geographers and other urban researchers as an explanation of past and present population changes in both urban cores and surrounding fringe areas. This model is useful to describe urban growth and decline periods in Europe through four sequential stages: urbanisation, suburbanisation, desurbanisation and reurbanisation, each one being sub-divided into two periods of relative or absolute population increase (centralisation) or decrease (decentralisation). In a first stage, urbanisation, the core city wins more population than the surrounding region, while the opposite occurs in the following phase, suburbanisation, when demographic decentralisation leads core cities lose relative demographic strength, while surrounding areas rapidly rise their inhabitant numbers. Though core cities still initially show population increase, growth first progressively slows down, and then finally starts diminishing. Even though in the suburbanisation stage all the urban area as a whole still shows positive growth, the sign shifts to negative in the subsequent phase, desurbanisation, when population decline appears both in core cities and fringe areas. While the latter ones start to present a negative demographic trend, central cities become increasingly de-populated (Hall, 2006). In extreme cases, these can even fall into disrepair, decrepitude and vacancy a phenomenon called urban decay by some authors like Medhurst and Parry Lewis (1969) or Andersen (2003), and shrinking cities by others (Oswalt, 2003; Oswalt and Rieniets, 2006; Ebers, 2007). When population and economic activities in rural areas and satellite non-metropolitan towns grow along with desurbanisation, the phenomenon has been called counterurbanisation (Fielding, 1982). Finally, desurbanisation should be followed by a fourth a final stage, reurbanisation, characterised by a progressive core city population recovery and a later fringe area decline rate reduction. Van den Berg

3 3 Workshop 04: Growing discrepancies between sustainability and affordability in the metropolitan housing markets in the period of public finance crisis et al. (1982) considered this fourth stage (which would complete and restart the urban development cycle) as purely hypothetical and unlikely. However population data collected in the 1990s and the early 21 st century shows that some core cities are once again gaining population and thus, reurbanisation is in fact taking place (Lever, 1993; Cheshire, 1995; Ogden and Hall, 2000; Haase et al., 2005; Buzar et al., 2007).This urban recovery phase is seemingly caused by the settlement in core cities of two parallel flows, that of local and that of foreign people. On the one hand autochthonous households (one-person and one-parent household, non-family shared households and other nontraditional ones) people, particularly youngsters, seeking better educational and work opportunities, or those just generally attracted by core cities due to their positional advantages and way of life (Champion, 2001b; Buzar et al., 2005; Kabisch and Haase, 2011) are reaching in urban centres. Such movements would be related to second demographic transition household structure and lifestyle changes (Van de Kaa, 1987; Lesthaeghe and Neels, 2002). On the other hand, foreign immigrants would also be moving there. These flows have been particularly strong in Spain and other Southern Europe countries with a highly segregated labour markets (Domingo and Gil-Alonso, 2007; Gil- Alonso and Domingo, 2008), which offer international immigrants low-paid jobs and favour their settlement in the core cities and particularly their low-quality neighbourhoods (Bayona and López, forthcoming). Meanwhile national citizens suburbanisation trends have continued growing (Bayona and Gil-Alonso, 2008; Bayona et al., forthcoming). This four-stage sequential model of urban development has been employed by several authors (Cheshire, 1995; Turok and Mikhnenko, 2007; Kabisch and Haase, 2011) to compare European urban trends. Our paper intends to apply the Kabisch and Haase (2011) adaptation of the Van den Berg et al. (1982) model to classify the 15 Spanish urban areas with more than 500,000 inhabitants. According to Nel.lo (2007), just before the present economic crisis, Spanish cities were undergoing the third stage of metropolitan development. After a first concentration period ( ) that is to say, and using the Van den Berg et al. (1982) model, the urbanisation phase, there was a strong dispersion stage, the so called suburbanisation period ( ), in which the main cities lost population (suburbanisation stage with absolute decentralization). From the latter date onwards, Spanish cities would have entered a third phase. Even though suburbanisation continued, urban metropolitan centres would have started to recover (suburbanisation stage with relative decentralisation). This core city demographic recovery and fringe area growth has also been observed by Kabisch and Haase (2011: 244) from 2001 onwards in the other southern European cities. However, the Spanish case would be somewhat different as it would not follow the usual recentralization model in which people return to the urban centre (Cheshire, 1995; Champion, 2001a). Those moving to metropolitan cores are in fact no longer Spanish people, who in most city centres are actually diminishing -though López-Gay, 2011, beliefs that this trend slowing down and that, in the near future, it will change sign. They would rather be foreign immigrants (see Bayona and Gil-Alonso, 2008; and Pujadas, 2009, for Barcelona; or Pozo and García, 2009, for Madrid) arriving during Spain s two decade long immigration boom (Muñoz and Izquierdo, 1989), and intensifying in the second half of the 1990 s. In Spain, this third phase concludes when the economic crisis irrupts, particularly striking the real estate market. For the purposes of this paper, we will therefore, consider it specifically lasting from January the 1 st 2000 to January the 1 st Consequently, a new stage started on this latter date. As we do not still have enough perspective, nor data (we only have final data for the period) nor a theoretical framework, it is too early to analyse or extract many conclusions from it. Maybe the most interesting recent contribution is has been made by Kabisch and Haase, who have applied the Van den Berg et al. theoretical model to European evidence obtained in these last two decades, reaching the conclusion that no consecutive order of the stages of urban development was identified. Rather, we would support the idea that the regular onward cycle of the stages of urban development (van den Berg et al.,1982) was proven to have been reversed due to the trends of reinforcing suburbanisation and developing reurbanisation after 2001 (Kabisch and Haase, 2011: 246). Therefore, this last decade, several stages coexisted while significant differences between European regions emerged: in Eastern European agglomerations, desurbanisation has been predominant, though

4 Mixité : an urban and housing issue? 4 in the other three regions -Northern, Western and Southern Europe- suburbanisation is still the dominant phenomenon and reurbanisation is also increasingly present. Using this theoretical framework, we intend to analyse this uncertainty period through its demographic changes, comparing it with the one before ( ) in which demographic growth was higher. Foreign and local populations will be analysed separately, as we suspect that, if it had not been for foreigner flows, after suburbanisation, Spanish urban areas would have probably entered -as the Van den Berg et al. (1982) model indicates- the desurbanisation one and that in these last two years some of the Spanish cities might have been driven towards it. Nevertheless, if we are to prove this hypothesis the two populations need to be analysed separately. By doing so, we will also be able to assess the stage at which each of the Spanish cities studied is situated and the impact of foreign immigration on each of them. Given the overall foreign migrant figures, these will undoubtedly be relevant. Foreign immigration in Spain Since the mid 1990 s, and particularly from the early 21st century, Spain has undergone a foreign immigration boom (Domingo, 2004; Izquierdo and López de Lera, 2003; Reques and De Cos, 2004), becoming the European country with the largest immigratory flows. In the year 2007, when immigration was at its peak, 920 thousand new foreigners were yearly registered. The year 2010, foreigners attained 12.2% of the population, that is to say, there were more than five million seven hundred thousand aliens registered 1. However, a decade before, foreigner figures did not even reach one million and did not represent more than 2.3% of the population. As it can be observed in figure 1, the number of foreign residents annually registered in the Padron is high, their figures basically changing according to its administrative variations 2. As a result of this constant foreign population growth, Spain became the European country with the highest non national population stock. Nevertheless, the described trend was abruptly interrupted by the outburst of the crisis. Spain has changed from a traditionally emigrational 3 to an immigational country (Muñoz and Izquierdo, 1989). The first foreign flows started to arrive in the late 1970 s. They consisted in low quantities of Moroccans who had increasing difficulties to enter other European countries and Latin-Americans, from very specific countries. Later, from the 1990 s onwards, these Moroccan flows started to grow, to which Latin-American ones firstly Peruvian, then, from 2000, Ecuadorian, and finally those from other countries in the region- added. Eastern European flows also progressively incorporated. In fact, from 2008, the foreign nationality with the highest number of residents in Spain was actually Rumanians. Finally, Asian flows should also be mentioned. Significant numbers of Communitarians, who in some Mediterranean and insular regions would mainly be retired people, should be also taken into account and added to all the latter mainly economic migrants. Thus, foreign population inhabiting Spain is extremely heterogeneous showing diverse national origins and age structures depending on the size of the municipality where they live and the main activity carried ot in that territory (figure 2). For example, while Asians are practically absent from municipalities under one thousand inhabitants, they tend to, where they have come to represent 12.4% of all the migrants living in large cities. In fact, one in three Asians can be found in Madrid or Barcelona. Latin- 1 Provisional padrón data for 2011, registering 17,067 foreigners less- therefore scarcely varies. However, the numbers of those born outside Spain increases in 55,704. Though it is an issue which cannot be checked, differences are surely due to naturalisations. 2 For example, the 2006 and 2007 supposed down fall in padrón registrations is most probably due to the fact that foreigners who do not have a permanent residence permit have to reregister every two years so that they can be eliminated from the register if they do no longer live in the country. Therefore, most of this population has left in former years. 3 According to the PERE -Padrón de Españoles Residentes en el Exterior- in 2011 there were still 1.7 million Spaniards living outside Spain -23.8% of which aged 65 or over.

5 5 Workshop 04: Growing discrepancies between sustainability and affordability in the metropolitan housing markets in the period of public finance crisis Americans, also mainly working in the tertiary sector, show similar patterns, being increasingly present as the size of the city increments (Vono, 2010; Gil-Alonso and Domingo, 2008). Africans, on their side, follow the opposite trends (only 6% of them live in Madrid or Barcelona). As for Europeans, they mostly also reside in small municipalities. In this case, the choice of the place of residence is either linked to their relationship with agriculture -agro-industry- or, to residential reasons. Figure 1. Foreign population change in percentages and foreigner annual growth according to the Padrón, % 14% 12% 10% 8% 6% 4% 2% 0% -2% , , , , , , , , ,000 Source: INE Padrón continuo. Annual population growth % foreigners Figure 2. Foreign population distribution by municipality size and origin, Spain More de 1 million inhabitants between 300,000 and 1 million between 100,000 and 300,000 between between 50,000 and 20,000 and 100,000 50,000 between 10,000 y 20,000 between 5,000 and 10,000 between Less than 1,000 and 1,000 5,000 inhabitants Africa Europe America Asia Source: INE 2010 Padrón continuo.

6 Mixité : an urban and housing issue? 6 In sum, immigration has mainly settled in large urban areas, the touristic seacoast or certain intensive agriculture ones. The percentage of foreigners in Barcelona and Madrid respectively reaches 17.5% and 17.4%, while that of those born abroad attains over 20%. However, in other cities like Seville (5.3%) or Cadiz (1.7%) they are much less significant and considerably under the Spanish mean. Despite being true that many large cities are situated above this mean and that, under the municipality level, they are concentrate in certain urban neighbourhoods (Bayona and López, forthcoming), cities are not the places with the highest foreigner percentages, Mediterranean coastal tourist municipalities are. In 31 of them, they represent more than half of the population, and, in some very extreme cases, they even are a 75% of it (basically, British and German people). On the opposite side there are 828 municipalities (10.4%) without any foreigners at all, this relatively low percentage indicating how widespread the phenomenon is. 6.7% of the immigrants even live in municipalities under inhabitants. Data and Urban Areas Definition Sources used in the paper are the 1991 population census from which population data has been extracted and employed as a starting point, and the most recent available Padrón continuo from January the 1st Population has been distinguished by nationality (foreign or local) while metropolitan areas have been divided into core and periphery. This stock data has been completed with municipal flow micro-data obtained from the Estadística de Variaciones Residenciales (EVR). Data on births and deaths from the Movimiento Natural de la Población (MNP) have also been included. Therefore, with this data obtained from the INE s web, we have been able to observe natural growth variations in metropolitan municipalities. The choice of the particular urban areas to be studied posed two main difficulties. On the one hand, there was the issue of where to situate the threshold from which each locality should or not be considered urban. And on the other, there was also the problem of geographical limits. The lack and difficulty of obtaining harmonized data, due to the absence of homogeneous official definitions of Spanish metropolitan areas -similar to MSA and CSA American ones- has led to the proposal of diverse delimitations, some of which the main ones- we include here: AUDES Áreas Urbanas de España ( those appearing in Nel lo (2004), Serrano (2007), Feria (2008, 2011) and those drawn in the Atlas del Ministerio de la Vivienda (2006). Nevertheless, none have been officially adopted. As there was no clear directive, for this study, we decided to employ the definition used in the Atlas de las Áreas urbanas de España (Ministerio de la Vivienda, 2006) and to situate the threshold at 500,000 inhabitants. In sum, fifteen metropolitan areas satisfied the requirements. The only exception to the former definition and limits has been Madrid, as the whole Autonomous Community has been taken into account. However, some authors like Pozo and García (2009) consider that this metropolis has even overgrown such limits. Even though the Atlas quite often uses administrative limits which do not always reflect population dynamics (the case of Madrid, where suburbanisation has overgrown its limits is not the only one), employing it increases result comparability with other studies using the same administrative definitions. Among the selected fifteen urban areas there are eleven which would correspond to the classical definition of a metropolitan area i.e. a core city giving it name and its metropolitan periphery. The other four, would have grown from a couple of centres. They are: the central urban area of Asturias (Gijón and Oviedo), Alicante (Alicante and Elche), Bahía de Cádiz (Jerez de la Frontera and Cádiz) Pontevedra (Vigo and Pontevedra). The number of municipalities composing these metropolitan areas ranges from four, in the case of Las Palmas de Gran Canaria, to 164 in the case of Barcelona and 178 in that of Madrid -it should be taken into account that the area analysed corresponds to the whole Autonomous Community. In all, they are 564 municipalities, representing 7.0% out of the present 8,114, holding more than 21 million inhabitants, that is so say, 46.6% of the Spanish population. The period has been divided into two: The first would go from 2000 to 2008, corresponding to the economic growth and large immigration flow years. In contrast, the last,

7 7 Workshop 04: Growing discrepancies between sustainability and affordability in the metropolitan housing markets in the period of public finance crisis including only two years, from 2008 to 2010, would be that of those pertaining to the present crisis and slow foreign population increases. Figure 3. Centre-periphery distribution of foreigners in the main Spanish metropolitan areas Source: INE 2010 Padrón continuo. Urban population growth Between 2000 and 2010, registered inhabitants increased from 40,499,790 to 47,021,031. As known, foreigners played a crucial role in the process, being approximately responsible for at least three quarters of the growth. Leaving naturalizations and children from mixed marriages aside, their numbers have multiplied by six, increasing from 923,879 to 5,747,734, or in other words, from 2.3% to 12.2% of the population, which is -it should be remembered- the highest absolute (six and a half million inhabitants more) and relative growth in a decade since the Table 1. Population change by nationality in the main Spanish metropolitan areas españoles extranjeros total españoles extranjeros total españoles extranjeros total Alicante-Elche (6) 535,488 25, , , , , , , ,039 Central de Asturias (18) 799,402 6, , ,486 33, , ,339 60, ,141 Palma de Mallorca (8) 408,141 22, , , , , , , ,550 Barcelona (164) 4,215, ,967 4,328,447 4,238, ,009 4,928,852 4,268, ,514 5,012,961 Bilbao (35) 898,720 7, , ,330 46, , ,255 53, ,085 Bahía de Cádiz (6) 578,907 3, , ,417 15, , ,799 17, ,172 Vigo-Pontevedra (14) 543,964 5, , ,070 25, , ,521 26, ,505 Granada (30) 427,731 4, , ,459 26, , ,665 28, ,815 Madrid (178) 5,033, ,487 5,198,622 5,266,257 1,005,381 6,271,638 5,378,740 1,079,944 6,458,684 Málaga (8) 712,001 40, , , , , , , ,919 Murcia (10) 497,736 6, , ,634 83, , ,482 89, ,043 Palmas Gran Canaria, L. (4) 484,169 12, , ,620 34, , ,225 35, ,088 Sevilla (24) 1,155,103 7,609 1,162,712 1,212,166 46,121 1,258,287 1,229,329 57,348 1,286,677 Valencia (45) 1,336,254 17,025 1,353,279 1,358, ,460 1,535,028 1,368, ,376 1,557,907 Zaragoza (14) 634,759 7, , ,764 85, , ,306 95, ,571 18,260, ,075 18,707,065 18,834,654 2,618,153 21,452,807 19,052,144 2,856,013 21,908,157 * Numbers between brackets correspond to the number of municipalities included in each metropolitan area. Source: INE 2000, 2008 and 2010 Padrón continuo.

8 Mixité : an urban and housing issue? 8 As a result, main metropolitan area inhabitants also grew. The 18,707,065 residents living in 564 municipalities studied in the year 2000, ten years later became 21,908,157 (table 1). These three million more dwellers represent more than half of the decade s Spanish demographic increase, the annual cumulative growth for all the metropolitan areas being 1.59% (1.43% for the rest of Spain). However, not all behave the same way. Even though population has raised everywhere, that of Madrid (2.19%) and the Mediterranean coast urban areas have particularly increased: 2.42% in Murcia; 2.39% in Palma de Mallorca; 2.28%; in Malaga or 2.19% in Alicante. As observed in figure 4 (where the highlighted axes show Spain s mean values), these urban areas are situated in the NE sector, meaning that the areas receiving more migrants are also those growing more. Figure 4. Relationship between annual cumulative growth (r%) and proportions of foreign population in the main Spanish metropolitan areas, Annual cumulative growth rate, R² = Source: INE 2000, 2008 and 2010 Padrón continuo. Proportion of foreigners in 2010 The less dynamic metropolitan areas have, on the contrary, received little immigration (SW sector), the smallest increases being: 4% for Bilbao; 0.37% for Área central de Asturias; 0.65% for Vigo- Pontevedra; 0.84% for Las Palmas de Gran Canaria and 0.93% for Bahía de Cádiz. The latter are in fact also those with the lowest percentages of foreigners (table 2). Granada would be the only exception. Despite having relatively low alien percentages, that is to say, under the Spanish mean, it has grown more than the countiy s as a whole. As it can be observed (figure 4), it is the only case situated in the NW sector. Barcelona, Zaragoza and Valencia, on their side, combine high foreigner percentages and growths under the Spanish population mean (SE sector). In sum, foreigners presence in urban areas is highly irregular, ranging between 21% (Palma de Mallorca and its metropolitan area) and 2.7% (Bahía de Cádiz). In fact, the proportion of immigrants living in seven of the metropolitan areas drops significantly under the Spanish mean (table 2).

9 9 Workshop 04: Growing discrepancies between sustainability and affordability in the metropolitan housing markets in the period of public finance crisis Table , 2008 and 2010 percentages of foreigners in the main Spanish metropolitan areas Palma de Mallorca Spain Madrid Central de Asturias Alicante-Elche Palmas de Gran Canaria, L Málaga Bilbao Barcelona Granada Murcia Vigo-Pontevedra Zaragoza Sevilla Valencia Bahía de Cádiz Source: INE 2000, 2008 and 2010 Padrón continuo. The impact of the economic crisis on demographic growth As expected, demographic growth is higher for the first ( ) period than for the second ( ), respectively increasing 1.73 and According to the number of registered residents, foreigners raise more rapidly than nationals. Moreover, in the crisis period, their annual growth even reached a 4.44% (table 3). However, despite being relatively significant, these figures are six times lower than those of previous years (24.76%). In other words, as shown by their plunging levels, foreigners received a hard impact. Oddly enough, as annual growth rates indicate, there were more local people living in large urban areas after 1/1/2008 than before that date (rising from an annual 0.39% to a 0.58%). This trend could possibly be explained by naturalisation figures rather than natural balances, as latter increases are so small that data obtained cannot otherwise be explained (Bayona et al., 2011). Table and population growth by nationality in the main Spanish metropolitan areas. Absolute figures and annual cumulative growth rates (r%) Spanish foreigners total Spanish foreigners total Sp. for. total Sp. for. total Alicante-Elche 41,293 83, ,684 6,049 5,236 11, Central de Asturias -4,916 27,176 22,260-19,147 27,000 7, Palma de Mallorca 18,362 83, ,962 4,738 8,153 12, Barcelona 23, , ,405 29,604 54,505 84, Bilbao -38,390 38, ,075 7,761 3, Bahía de Cádiz 36,510 11,882 48,392 6,382 1,964 8, Vigo-Pontevedra 11,106 19,221 30,327 4,451 1,951 6, Granada 40,728 22,009 62,737 15,206 1,810 17, Madrid 233, ,894 1,073, ,483 74, , Málaga 68,755 94, ,213 13,678 13,526 27, Murcia 41,898 76, ,685 11,848 5,976 17, Palmas Gran Canaria, L. 16,451 21,703 38,154 3,605 1,575 5, Sevilla 57,063 38,512 95,575 17,163 11,227 28, Valencia 22, , ,749 9,963 12,916 22, Zaragoza 6,005 78,401 84,406 5,542 9,697 15, ,664 2,172,078 2,745, , , , Source: INE 2000, 2008 and 2010 Padrón continuo. As for precise metropolitan areas, that of Bilbao is odd case, as between 2000 and 2008 it had a stagnant population (only increasing by 177 inhabitants), and then, in the following two years, it gained 3,686 residents. Though, between those dates, none of the areas studied lost absolute population, they all won less than in the preceding period. However, to compare them properly, annual

10 Mixité : an urban and housing issue? 10 cumulative growth rates should be taken into account. From this perspective, the área central de Asturias (increasing from a 0.34% in the first to a 0.47% in the latter period) and Seville (rising from 0.99% to 1.12%) should be added to Bilbao (making, in , after not growing at all in the previous period (0% in ) a small increase of 0.20%). On their side, in the latter period, Vigo-Pontevedra and Granada, very slightly reduced their rates. As these are the five cities where foreign immigration has been relatively less present (table 2) they are probably also those less affected by the economic crisis. On the contrary, growth rates of metropolitan areas with the highest proportions of foreigners have considerably fallen. Centre periphery differences Demographic relevance of each of the 15 central cities with respect the rest of their own metropolitan area differs considerably. From Barcelona, containing only 23% of the metropolitan residents to Zaragoza holding 91 % of them, there would be a whole number of intermediate situations. Two elements play a crucial role in these differences, the number of municipalities considered and the central city s size in spatial terms-, the latter ranging from 12 km 2 in Cadiz to km 2 in Madrid. Globally speaking, however, during this decade, population rose and expanded that is to say, it increased more in the periphery than in central cities. As it can be observed on the following maps (figure 5, absolute figures, figure 6 relative ones), in 12 out of the 15 cases analysed, both core cities and fringe areas have grown. Figure centre and total Spanish metropolitan area absolute growth. Source: INE 2000 and 2010 Padrón continuo. The three exceptions to this pattern would be the three metropolitan areas furthest on the right (figure 6). While both central cities of Granada and Bilbao lose population, in the area central de Asturias, the centre grows but its periphery diminishes. In sum, except for the latter case, which would still be at the initial urbanisation stage where there is absolute centralization (Van den Berg et al., 1982), the rest would be at a deconcentration phase, which would be absolute (Granada and Bilbao) or relative (the rest of metropolitan areas) depending on the cases. Malaga where the people living in the core city fall from 73.8% in 2000 to 60.3% in and Granada - where it dropped from 56.4% to 46.7%- would be the most paradigmatic deconcentration cases. As formerly indicated, the latter city s residents diminish by 0.22% while its periphery grows 3.81%. However, the case of Bilbao is slightly different as its population distribution remains stagnant, its central municipality only increasing by 3% and its periphery by 9%.

11 11 Workshop 04: Growing discrepancies between sustainability and affordability in the metropolitan housing markets in the period of public finance crisis Figure centre and periphery Spanish metropolitan area annual cumulative growth rate Centre Periphery Metropolitan area Source: INE 2000 and 2010 Padrón continuo. Table Centre and periphery Spanish metropolitan areas natural and migratory growth. Total Natural Net Total Natural Net Total Natural Net Growth Change Migration Growth Change Migration Growth Change Migration Centre 92,563 19,766 72,797 Centre 10,984 8,652 2,332 Centre 84,179 23,198 60,981 Periphery 43,406 5,726 37,680 Periphery 45,754 18,933 26,821 Periphery 52,330 13,103 39,227 Alicante- 135,969 25, ,477 Bahía Cádiz 56,738 27,585 29,153 Murcia 136,509 36, ,208 Elche Centre 34,516-14,158 48,674 Centre 18,367 3,569 14,798 Centre 24,790 8,428 16,362 Periphery -4,403-14,160 9,757 Periphery 18,362 3,363 14,999 Periphery 18,544 6,584 11,960 Central de Asturias 30,113-28,318 58,431 Vigo- Pontevedra 36,729 6,932 29,797 Palmas G. Canaria, L. 43,334 15,012 28,322 Centre 70,756 13,179 57,577 Centre -5,332 2,726-8,058 Centre 3,482 18,382-14,900 Periphery 44,097 6,614 37,483 Periphery 85,085 17,289 67,796 Periphery 120,483 37,799 82,684 Palma de 114,853 19,793 95,060 Granada 79,753 20,015 59,738 Sevilla 123,965 56,181 67,784 Mallorca Centre 123,071-19, ,177 Centre 390,189 59, ,914 Centre 70,253 8,508 61,745 Periphery 561, , ,572 Periphery 869, , ,761 Periphery 134,375 28, ,348 Barcelona 684, , ,749 Madrid 1,260, , ,675 Valencia 204,628 36, ,093 Centre -1,084-6,669 5,585 Centre 36,942 17,312 19,630 Centre 70,490 5,112 65,378 Periphery 4, ,368 Periphery 153,475 18, ,320 Periphery 29,155 3,180 25,975 Bilbao 3,863-6,090 9,953 Málaga 190,417 35, ,950 Zaragoza 99,645 8,292 91,353 Source: INE 2000 and 2010 Padrón continuo and Movimiento Natural de la Población From the year 2000 onwards, most of the demographic growth is due to foreign immigration. These have not only strengthened both core city and fringe area positive migratory balances, but, to a certain stage, allowed natural balances births minus deaths- to recover. The only exceptions would be the cities of Seville and Granada which would be losing population due to emigration. Their peripheries would however be receiving high volumes of new residents. The impact of the arrival of these large amounts of immigrants was twofold; on the one hand they provoked a birth rise and, on the other, they slowed ageing down. Except for three cases (table 4), both core city and fringe area natural balances are positive. Though it is true that Barcelona and Bilbao peripheries have positive natural balances, their central cities do not, the only metropolitan area where both areas lose residents due to deaths and lack of births being the central de Asturias. In fact, as its periphery s migratory balance is positive though not large enough, it cannot even recover its population this way.

12 Mixité : an urban and housing issue? 12 Considering Spanish population on its own (fig 7, left) trends are quite different. Both the centre and the periphery of the area central de Asturias and Bilbao lose nationals and only five metropolitan centres (Murcia, Alicante-Elche, Jerez-Cádiz, Las Palmas and Vigo-Pontevedra) gain local residents even though not as much as their peripheries. In eight cases, including the largest metropolitan areas, fringe area locals increase and particularly intensely in Zaragoza, Málaga, Granada- while their core cities show negative balances. Within this group, Barcelona and Madrid would particularly show up due to their relevance. On the other hand, in figure 7 (right) it can be observed how foreigners highly and widely grow. Nevertheless, their settlement trend is somewhat obscure. While in Granada, Zaragoza, Alicante-Elche, Madrid, Valencia or Barcelona they seem to move to the periphery, in Málaga, Palma de Mallorca o la Bahía de Cádiz they seem to prefer the centre. Figure centre and periphery Spanish metropolitan area cumulative growth rate for locals and foreigners. Spanish Population Foreign Population Centre Periphery Metropolitan area Source: INE 2000 and 2010 Padrón continuo. Centre Periphery Metropolitan area In any case, foreigners relevance and number has not stopped increasing since the year As shown in table 5, despite the crisis, urban centres and peripheries have not stopped receiving them. However, it should also be noted that their increase is no longer as strong. Malaga s periphery has the highest proportion of metropolitan foreigners (27.5%) while Las Palmas de Gran Canarias fringe area has the lowest (3.2%) periphery proportions. On the other hand, Palma de Mallorca s (20.7%), Barcelona s (17.5%) and Madrid s (17.4%) urban centres have the highest core city proportions and the two Bahía de Cádiz (2.2%) centres, the lowest. Figure 8 allows differentiating national and foreign population core city and fringe area economic growth and crisis periods. Except for Granada, Bilbao and Seville all other core cities grew, between 2000 and However, all the peripheries except for the Área central de Asturias where growth was negative- increased even more, being the cases of Zaragoza and Malaga, particularly significant. This generalised urban centre expansion cannot be understood without the enormous foreign immigration inflows they received -8 of them even losing Spanish population. The cases of Granada (- 1.04% annually) and Barcelona (-0.96% annually) would especially stand out. On the other hand, except for the central de Asturias (-0.47%) and Bilbao (-0.38%), all the other peripheries gain local residents.

13 13 Workshop 04: Growing discrepancies between sustainability and affordability in the metropolitan housing markets in the period of public finance crisis Figure and centre and periphery foreign, Spanish and total annual cumulative growth rate for Spanish metropolitan areas Total Population Total Population Centre Periphery Metropolitan area Centre Periphery Metropolitan area Spanish Population Spanish Population Centre Periphery Metropolitan area Centre Periphery Metropolitan area Foreign Population Foreign Population Centre Periphery Metropolitan area Centre Periphery Metropolitan area Source: INE 2000, 2008 and 2010 Padrón continuo.

14 Mixité : an urban and housing issue? 14 Interestingly enough, the crisis seems to have particularly affected peripheries and those centres which grew more during the previous period. By contrast, core cities receiving less impact during growth years are also those less modified by the crisis. Seville and Granada even show small positive growths. This diversity of impacts seems to be due to the large immigration reduction to which metropolitan areas are submitted as a consequence of the crisis. While in the city of Granada flows even become negative. At the opposite end we can find the central de Asturias urban cores where immigration has largely increased. Probably driven by demographic inertia, local population seems to be less affected by the crisis. The number of urban centres losing Spaniards even reduces from 8 to 6. Granada, Madrid, Palma de Mallorca and Zaragoza, which had reached that situation, started gaining them. On the contrary Malaga, and particularly the two cities forming the central de Asturias, are those having greater falls. In Barcelona, Valencia, Seville and Bilbao, on their side, Spanish residents diminish in both periods, though less in the second. Except for Alicante-Elche, the rest gain it during both periods, though more in the latter. Table , 2008 and 2010 Centre and periphery Spanish metropolitan area foreigner percentages Centre 4,7 16,0 16,5 Centre 0,4 1,9 2,2 Centre 1,3 13,0 13,7 Periphery 3,9 15,6 16,1 Periphery 0,9 3,0 3,3 Periphery 1,5 14,2 14,5 Alicante-Elche 4,6 15,9 16,4 Bahía Cádiz 0,6 2,4 2,7 Murcia 1,3 13,4 14,0 Centre 1,0 5,1 8,9 Centre 1,2 5,1 5,3 Centre 2,9 7,8 8,0 Periphery 0,6 2,5 4,8 Periphery 0,8 2,9 3,3 Periphery 1,5 3,0 3,2 Central de Asturias 0,8 4,1 7,3 Vigo-Pontevedra 1,1 4,3 4,6 Palmas G. Canaria, L. 2,5 6,4 6,6 Centre 3,9 19,5 20,7 Centre 1,3 6,4 6,3 Centre 0,7 4,3 5,3 Periphery 9,9 21,2 21,7 Periphery 0,6 4,3 4,8 Periphery 0,6 2,9 3,4 Palma de Mallorca 5,2 19,9 21,0 Granada 1,0 5,3 5,5 Sevilla 0,7 3,7 4,5 Centre 3,1 16,9 17,5 Centre 3,4 16,8 17,4 Centre 1,6 14,2 14,8 Periphery 2,4 12,6 13,6 Periphery 2,9 15,2 16,1 Periphery 0,9 8,5 9,3 Barcelona 2,6 14,0 14,9 Madrid 3,2 16,0 16,7 Valencia 1,3 11,5 12,2 Centre 1,1 7,1 8,0 Centre 1,2 7,1 8,0 Centre 1,1 11,9 13,0 Periphery 0,6 3,8 4,6 Periphery 15,5 27,0 27,5 Periphery 1,0 10,4 11,3 Bilbao 0,8 5,1 5,9 Málaga 5,4 14,7 15,7 Zaragoza 1,1 11,8 12,8 Source: INE 2000, 2008 and 2010 Padrón continuo growth of the main Spanish urban areas: a typology As a final contribution to the analysis, the 15 metropolitan areas have been classified according to a series of centre and periphery growth indicators extracting a typology. As the number of variable analysed was high, a cluster was carried out. Six groups, with high explanatory power have come out. Even though originally we intended to have fewer categories, the metropolitan area variety and clear differences between cities pertaining to smaller groups and those in the rest of categories led us to maintain results even though one category contains one only urban area and three hold two (table 6). The predominant urban features at the fifteen urban areas analysed belong to suburbanisation with relative decentralisation stage. Even though urban centres have gained population (an annual mean of 0.96%) suburban municipalities show the highest growths (2.31%). Therefore, from the relative point of view, core cities have lost population. In 2010, the percentage of people living in central cities was 3.3% less than in the year 2000, decreasing from 54.7% to 51.4%. This trend is highly related with the large amount of migrants which have arrived. Despite preferably settling, at least in a first stage, in core cities (13.6% of their population is foreign) and less in peripheries (12.5%) they have reinforced Spanish population s suburbanisation dynamics. More specifically, young Spaniards migration and settlement behaviour are the main causes of larger periphery migratory growth (17.3 per thousand

15 15 Workshop 04: Growing discrepancies between sustainability and affordability in the metropolitan housing markets in the period of public finance crisis compared to 8.2 per thousand in core cities, basically gaining foreigners) but also of their natural growth (5.4 per thousand in comparison to 1.4 per thousand in core cities). Even though peripheries only contain 48.6% of the population, they produce 77.9% of the natural growth and 65.4% of the migratory one. Table 6. Main Spanish metropolitan area typology according to diverse growth indicators % change Annual growth rate (%) % foreig people Natural growth rate ( )Migratory growth rate ( ) people living in core cities Core city Periphery Core city Periphery Core city Periphery Core city Periphery Category Category Category Category Category Category TOTAL Source: INE 2000 and 2010 Padrón continuo and Movimiento Natural de la Población However, there obviously are differences among them. Both the periphery and core cities of those pertaining to category 1 (Alicante-Elche, Madrid, Murcia and Palma de Mallorca) have substantial amounts of foreigners, always above the Spanish mean. Even though they are all undergoing a strong suburbanisation process and the central city is therefore losing force, the metropolitan core is still considerably increasing. As both their centres and suburban areas show high natural and migratory balances, these metropolises are the most demographically dynamic. They are also the locations where large amounts of migrants choose to settle. Madrid has a large power of attraction due to its status as capital and has become a migrant s entry door; Palma de Mallorca and Alicante have a large amount of Communitarian residents and Murcia concentrates on intensive agriculture. Despite having similar characteristics to those of the former group, category 2 urban areas, that is to say Valencia and Barcelona, have grown less. Their core cities are less dynamic and their suburbanisation flows smaller. Foreigner percentages in the central city continue being high, but they are slightly lower in fringe areas. Even though natural growth is even negative in the city of Barcelona, with immigration, both cities recover population, due to high migration balances. This would be particularly significant as, in the previous period, they had absolute losses. The main characteristics of category 3 metropolitan areas are their extremely strong suburbanization process, and that their core cities are subsequently becoming less significant from the demographic point of view. In effect, Málaga s and Zaragoza s peripheries are growing more than their core cities due to the extraordinarily high migratory balances combined with a significant natural growth. This is due to their particularly young population structure, a characteristic of this young household migration towards the periphery. In both metropolitan areas, the percentage of foreigners is similar to that of Spain as a whole. Finally, it should be stressed that in Malaga, they particularly settle in fringe areas, a characteristic which should be related to the fact that the main type of migrants moving there are Communitarians. Category 4 is only made up of one metropolitan area, Granada. The main reason behind it is that, during the analysed decade, its core city relative population has dropped by 10%. Despite having positive natural growths central city loses population due to the small flows arriving and their subsequent negative migratory balances. Peripheries, on the contrary, show the opposite trends, as they are quickly suburbanising due to high Spanish immigration flows and show one of the highest natural growth rates. In summary, the main difference with the rest of high suburbanised cities is that the lack of foreign migrants is leading to a core city population loss.

16 Mixité : an urban and housing issue? 16 Foreign population percentages in Bahía de Cádiz, Las Palmas de Gran Canaria, Sevilla and Vigo- Pontevedra core cities and fringe areas, forming category 5, are extraordinarily low, very much under the Spanish mean, and in the periphery particularly so. These cities are the only ones where natural growth is higher than migratory balances. Bahía de Cádiz and Seville are two southern Spain metropolitan areas where international migration s incidence is low and which are showing similar trends to those of the North Western metropolis of Vigo-Pontevedra. However, the case of Las Palmas de Gran Canaria is somewhat different. Even though there are large amounts of foreigners in the island, they do not live in its main city. These mainly Communitarian residents prefer to do it in the touristic areas. Even though they have not received much immigration, suburbanisation is also present. Despite their core cities have, in absolute figures, gained a certain amount of population, this is not true from the relative point of view. Finally, the two last metropolitan areas Bilbao and Central de Asturias belong to category 6, being, their stagnant population and the lowest suburbanisation flows, their main characteristics. In fact, Gijón and Oviedo, the central de Asturias core cities, would be recovering population while their fringe areas lose it. Therefore, this would mean that they are still at the urbanisation stage. On the other hand, the population of Bilbao s metropolitan area is stagnant, as it slightly diminishes in the centre and hardly grows in the periphery. Natural growth for both central cities is clearly negative and for Asturias fringe area, extremely so. On its side, that of Bilbao is practically nonexistent. In sum, despite 8% of these core cities population is foreign, their general demographic trend is no different from that of the rest of northern Spain. Discussion and conclusions In the early 21 st century and as a consequence of the massive arrival of foreign migrants, all the metropolitan areas with more than half a million residents have gained population. However, growth magnitudes have been extremely different. On one extreme we find the metropolis of Bilbao, where population has maintained nearly stagnant, and on the other, that of the Comunidad de Madrid, which has increased by about 1,250,000 inhabitants. Local population trends naturalisations should also be taken into account- indicate that, without immigrants, urban growth would have been much lower or in the case of Bilbao and the area central de Asturias have even been negative. The only metropolis still undergoing urbanisation and absolute centralisation is precisely the latter. The rest would either be experiencing suburbanisation within the absolute phase (Bilbao and Seville and Granada before 2008) or relative (the rest) decentralisation. However, this typology is influenced by the large amount of immigrants which have arrived to the country. If only local population developments are taken into account, the situation becomes quite different. Then, the four main cities - Madrid, Barcelona, Valencia and Seville- plus Zaragoza, Palma and Granada would also be under absolute decentralisation. However, as a consequence of the crisis, they seem to be directly initiating reurbanisation without going through desurbanisation. The effects of the crisis and the constant Spanish population loss have also seemingly led area central de Asturias to this latter mentioned stage -without going through suburbanization. Bilbao would also be undergoing desurbanisation, though it is showing recovery signs. These results seem to be in the line with those found by Kabisch and Haase (2011) for Europe as a whole. According to these latter authors, the Van den Berg et al. (1982) urbanisation stages would not follow one another but would coincide in time. Therefore, at one moment in time, different cities would be situated at diverse stages depending on their morphology, size and economic activities developed in them. Finally, answering the question set at the title, the economic and real estate crisis seems in effect to have restrained suburbanization dynamics. On the one hand, there are many less foreigners arriving, so metropolitan areas have reduced their rhythm of growth. On the other, as getting access to new dwellings has become more difficult, Spanish population is moving less and exit flows have been