Received: 15 July 2020 Revised: 17 December 2020 Accepted: 28 January 2021 DOI: 10.1002/psp.2446 RESEARCH ARTICLE School segregation of migrants and their descendants in a dual school system: The case of Barcelona Jordi Bayona-i-Carrasco1,2 | Andreu Domingo2 1 Department of Geography, University of Barcelona, Barcelona, Spain Abstract 2 Centre for Demographic Studies/CERCA, This study analyses levels of school segregation of students of immigrant origin in a Catalonia, Spain city of the south of Europe, namely, Barcelona, which is characterised by a rapidly Correspondence increasing growth of international immigration in recent decades and moderate or Jordi Bayona-i-Carrasco, Serra Hunter Fellow, even low levels of residential segregation of immigrants. Besides its analysis of Department of Geography, University of Barcelona, Barcelona, Spain, and Centre for nationality and origin, its main contribution is that it explores a generational typology Demographic Studies/CERCA, Catalonia, classifying students according to their place of birth, year of arrival and origin of pro- Spain. Email: jordibayona@ub.edu genitors, thus revealing different degrees of school segregation in keeping with each student's migratory history. Moreover, in the context of a dual school system (public Funding information Ministry of Economy and Competitiveness, and private), segregation is analysed by breaking it down in accordance with state or Grant/Award Number: CSO2017-85670-R private ownership of the school. The results indicate a marked degree of segregation among students of the first generation and also among descendants of immigrants, showing how the dual school system is responsible for a significant part of the segre- gation, and that comparison of residential and school segregation always shows that the latter is greater. KEYWORDS descendants of immigrants, immigration, residential segregation, school segregation, school system 1 | I N T RO DU CT I O N subsequent dispersal around the country, the numbers have always been higher, to such an extent that, in Barcelona today, one in four of The intensity of international immigration to Spain in the early years the city's residents was born abroad. of the 21st century with what has been called the international migra- In the case of Barcelona, the locus of this study, the territorial dis- tory boom (Arango, 2009; Domingo & Cabré, 2015) brought about tribution of foreign immigrants in the neighbourhoods of the city, is structural changes in Spanish society, with the incorporation within a characterised, as in other Spanish cities, by initial settlement in the old few years of five million immigrants, who came to represent 14.2% of centre, followed by their later swift distribution through all the city's the population. This affected all sectors of the state, including schools neighbourhoods, especially those in the periphery, which were con- where the impact was particularly noticeable. The arrival of immigrant structed in the 1960s and 1970s in response to Spain's earlier internal minors with their parents in the early stages of the process and subse- migratory processes. Only the more prosperous neighbourhoods, quent family reunification meant a multiplying presence of immigrant together with some of the most recent private housing estates, con- students in schools and the progressive incorporation, some years tinue to show moderate percentages of immigrants. In this setting, later, of their descendants, the so-called second generations (García- notable for its considerable dispersion, the values for residential seg- Castaño & Carrasco, 2011; González-Ferrer & Cebolla-Boado, 2018; regation are generally moderate or even low for most groups of immi- Portes et al., 2016), some of whom retained the nationality of their grants (Bayona & López, 2011; Galeano & Bayona, 2018; Martori parents. In cities like Barcelona or Madrid, which act as a gateway of et al., 2006), despite a wide range of situations depending on national entry for international immigration and also as points of its origins. Popul Space Place. 2021;e2446. wileyonlinelibrary.com/journal/psp © 2021 John Wiley & Sons, Ltd. 1 of 14 https://doi.org/10.1002/psp.2446 2 of 14 BAYONA-I-CARRASCO AND DOMINGO However, what happens in the residential domain does not existing between the various generational categories and the autoch- always match with the actual distance between immigrants and thonous population. In doing so, we first give an account of the immi- autochthonous inhabitants. In cities of the south of Europe, segrega- grant population in Barcelona, focussing on presence in schools and tion values tend to be moderate or low, even though they coincide employing as variables in the analysis nationality and place of birth. with high levels of vulnerability among immigrants (Arbaci, 2019; Second, we use a generational typology and present the results for Arbaci & Malheiros, 2010; Martínez Del Olmo & Leal, 2008). A clear different school years. Third, we apply segregation indicators with example of these discrepancies appears with analysis of what happens three different aims: (1) offering a comparison between school segre- in schools. In the case of Barcelona, there is a significant degree of gation and residential segregation, (2) calculating segregation for the overrepresentation of immigrant students in public schools but also different generational categories and (3) breaking down segregation in marked differences between schools in the public education system, a keeping with type of school (public and private), which allows us to situation that has been condemned by the Síndic de Greuges evaluate the influence of the dual school system on segregation. 1 in several of its reports (Síndic de Greuges, 2008, 2016). Hence, Finally, the implications of this analytical perspective on segregation school segregation is greater than residential segregation and its results will be discussed in the conclusions. (Alegre, 2017; Bonal et al., 2019; López-Falcón & Bayona, 2012), and this is particularly noticeable among the more residentially dispersed groups like those from Latin America (López-Falcón & Bayona, 2012). 2 | THEORETICAL FRAMEWORK AND Recently, in 2019, the Barcelona City Council announced its first pro- STATE OF THE ART gramme to combat this segregation by distributing 2,000 vulnerable students to schools with more resources, providing financial help and Although most studies of segregation use the residential sphere, there establishing new criteria for the management of ‘live enrolment’ of are other areas of individuals' lives where segregation can be even immigrant students joining their classes in the middle of the school more important than in the place of residence and, accordingly, have a year, which has previously been highly concentrated. greater impact on the process of adaptation to the host society. These Studying school segregation of foreign-born students, usually car- include schools in the case of children, the workplace for employed ried out on the basis of nationality, produces biased results. It renders people, the place where shopping is done, the means of transport invisible students who acquire Spanish nationality, many of them from used and leisure spaces, all of which complement the residential per- Latin America, while descendants of groups for which access to spective as places where interaction between immigrants and autoch- nationality is more difficult (e.g., Africans and Asians) keep appearing thonous residents takes place (Boterman & Musterd, 2016). These as foreigners. For example, 47% of students of foreign nationality in areas make up a very important part in the lives of individuals, and we the city of Barcelona were born in Spain while, among those students could even attribute greater significance to them than place of born abroad, 35.4% have Spanish nationality. In order to remedy this residence—because they play a key role in the possibility for interac- bias, which involves unequal persistence or disappearance in student tion with the host society—when referring to the development of the statistics depending on the criterion of observation, the present study process of integration. Studies that explore distance from autochtho- adopts a generational perspective in its calculations of segregation, nous residents in these areas identify a range of relationships this being based on the student's place of birth, year of arrival and between different types of segregation. Hence, school segregation birthplaces of the parents. This typology offers a fresh way of ana- is usually greater than residential segregation (Gramberg, 1998; lysing school segregation, contributing a new, all-inclusive view of Schindler, 2007), although the two types tend to be related segregation of students of immigrant origin. In Spain, very few studies (Boterman, 2019; Johnston et al., 2004). However, labour segregation include the student's birthplace. Most statistical series include the cri- czak et al., 2015). In the particular case of school is much less (Marcin terion of nationality (Spanish or foreign, noting the applicable nation- segregation, our concern here, the interest of our analysis stems from ality in the latter case), although the availability of statistical data in the fact that this is one of the key mechanisms for understanding the open access is very limited. The only exception is work drawing on reproduction of inequalities and possibilities for social mobility among the PISA (Programme for International Student Assessment) report in immigrant students (Boterman et al., 2019), because segregation has a which segregation is calculated using the category of immigrant stu- negative impact on the most disadvantaged students (Bonal & dents on the Autonomous Community scale (Murillo et al., 2018), Bellei, 2018). In a context of growing urban segregation in Europe starting out from student samples. Use of ethnic categories, as hap- (Musterd et al., 2017; Tammaru et al., 2016), Bonal and Bellei (2018) pens in other countries (Johnston et al., 2004), identifies a greater show how school segregation is once again attracting considerable number of students but does not supply data on the year of arrival. attention, as is evident in the appearance of recent research initiatives Besides explaining the process of integration of minors of immi- in several European cities. grant origin, analysis of school segregation offers a complementary The school system is one of the key explanatory variables with viewpoint that helps to achieve a more accurate interpretation of the regard to segregation. In Spain, as in other countries, there exists a process of integration of the immigrant population in domains other dual, public and private, educational system, although the latter, with than residence. We work with the hypothesis that there is a hierarchy a substantial presence, especially in the city of Barcelona, mostly takes in the values of school segregation determined by the distance the form of publicly subsidised private schools.2 Given this situation, BAYONA-I-CARRASCO AND DOMINGO 3 of 14 the possibility for parents of choosing a school is a determining factor of Barcelona, the economic crisis has had a major impact and the more in the growth of segregation (Alegre, 2017; Wilson & Bridge, 2019). In vulnerable zones are now to be found in the peripheral areas (Sarasa Catalonia, the initial choice of school is, according to the Catalan et al., 2018), which have a growing immigrant presence at the same Ombudsman's report (Síndic de Greuges, 2016), one of the main cau- time as segregation is increasing (Rubiales, 2020; Sorando & ses of school segregation. Moreover, progressive diversification of the Leal, 2019). In these vulnerable neighbourhoods, the concentration of educational projects of primary and secondary schools makes it easier the immigrant population in schools is normally greater (Johnston for autochthonous families, which are more informed of the charac- et al., 2007; Robinson, 1984), owing to a higher birth rate, an ageing teristics of and differences between centres, to make the selection. demographic structure in the autochthonous population and other This is especially true of middle-class families, as Benito and factors of choice, both positive (concentration of students of similar Gonzàlez (2007) note when speaking of the case of Catalonia. In this origins) and negative (fruit of discrimination). sense, and in relation with the immigrant population, segregation is In the case of Spain, the high levels of school segregation have, supported by two kinds of practices of the local population. One has according to Murillo and Martínez-Garrido (2018) who draw on data been called ‘ethnic avoidance’ (Rathelot & Safi, 2014), whereby par- of the PISA report, come to be among the uppermost on the ents decide against schools with larger proportions of immigrant stu- European scale. This would have intensified recently as a diffuse dents, and the other is ‘white flight’ where their children are taken effect of the economic crisis (Murillo & Martínez-Garrido, 2018). out of schools with larger proportions of immigrant students School segregation is receiving increasing attention in Catalonia (Sánchez-Hugalde, 2009) and sent to a private school or another pub- (Alegre, Benito, & González, 2008; Benito & Gonzàlez, 2007; lic school with a smaller proportion of foreign students. The impact of Bonal, 2012; Bonal et al., 2019; Carrasco et al., 2007; Ferrer the economic crisis, in which countries like Spain are among the et al., 2011; Sánchez-Hugalde, 2009; Tarabini et al., 2018), with par- hardest hit, only aggravates these differences. ticular interest from the institutions (Síndic de Greuges, 2016). This Other factors also influence school segregation. The most impor- school segregation in Catalonia is greater than residential segregation tant of these is socioeconomic stratification in the host society, where (Alegre, 2017; Bonal et al., 2020), especially in the case of students students are segregated in accordance with the social characteristics with Latin American nationalities, where the low levels of segregation of their families of origin (Jenkins et al., 2008; Murillo et al., 2018). from the residential point of view are not the case when it comes to Hence, foreign students are segregated as members of the most schools (López-Falcón & Bayona, 2012). underprivileged social classes, because immigrant families with Analysis of school segregation and its prevention is important school-going children tend to be in situations of disadvantage. This because of the effects involved. In particular, high levels of segrega- process is favoured in the context of the dual educational system with tion mean less contact with local students, which, in turn, affects the public and private schools. On the one hand, both systems are process of integration (Johnston et al., 2007) and school results. unequally distributed throughout the territory (Bonal & Indeed, evidence from Spain demonstrates that high levels of concen- Zancajo, 2020; Bonal et al., 2020) and, notably, with a smaller pres- tration affect school results. Data from the PISA report show that, as ence of private schools in vulnerable areas, as happens in Barcelona the number of immigrant students increases, the threshold at which (Domingo & Bayona, 2021). On the other hand, it encourages segre- the effect is observed also rises. Cebolla-Boado and Garrido (2011), gation because of the possibility of choosing among schools and the using data from the 2006 PISA report, found negative effects when higher economic costs associated with private schools, even though the threshold exceeded 20%, while Calero and Escardíbul (2016) con- they might be partially state-funded. In these cases, school zoning pol- clude from the 2012 PISA report that the threshold increases to 30% icy and choice of school are among the factors that have an effect for autochthonous students and 40% for foreigners. Even when it is (Alegre, 2017; Oberti & Savina, 2019), either by facilitating or of low intensity, concentration has negative effects for students of preventing segregation. In the case of Catalonia, the non-existence, both immigrant and local origins (Bayona & Domingo, 2019). for many years, of clear policies in this regard has ended up encourag- Nordin (2013) believes that the negative effects of concentration on ing segregation between the systems (Bonal, 2012). This is what has school results are selective, especially affecting boys and second- happened in Barcelona where school zoning reforms favouring greater generation students. eligibility have brought about increased segregation (Bonal & Analysis of segregation from the generational perspective, which Zancajo, 2020). is one of the focusses of the present study, is essential for discovering At the same time, segregation of immigrant students is also the impact of migratory trajectories on the distribution of immigrant explained by isolation of immigrant groups in the city (Bonal students, as well as for observing changes in segregation over time et al., 2019; Johnston et al., 2004; Schindler, 2007), and its evolution (Fiel & Zhang, 2018). For immigrants as a whole, the theory of spatial over time might be related with changes in the composition of the assimilation indicates that descendants of immigrants show greater immigrant population (Johnston et al., 2008), in which case the school dispersion than the earlier generations (Allen & Turner, 1996; White, is a reflection of what is happening in society. Thanks to expanding Biddlecom, & Guo, 1993) and, accordingly, lower residential segrega- globalisation and the economic crisis, Europe has recently seen a tion. The children of immigrants, the second generation, should have growth of more vulnerable zones, as a reflection of increasing social fewer difficulties because of their better knowledge of the language inequalities (Musterd et al., 2017; Tammaru et al., 2016). In the case and the educational system by comparison with the first generation. 4 of 14 BAYONA-I-CARRASCO AND DOMINGO In Spain, Aparicio and Portes (2014), working with the second genera- by the parents' year of arrival and place of origin, are important fac- tion, identify the existence of good school results despite a high risk tors in the process of adaptation to the host society. of failure for students of certain origins. This differentiation by origins has also been found by Bayona and Domingo (2018) when studying the school results of the second generation, in which Sub-Saharan 3 | DA T A A N D M E T H O D O L O G Y and other African groups are at the lower end of the scale, a situation which De Miguel and Solana (2017) also find for students coming Student microdata from official nonuniversity studies carried out in from North African countries. This lesser degree of school segregation the city of Barcelona, ceded by the Department of Education of the is not always fulfilled because, in some cases, a greater prevalence of Generalitat (Government) of Catalonia have been used. The data are segregation has been found among the second generations from the 2015–2016 school year and cover the totality of non- (Schindler, 2007). university enrolled school children in Catalonia, which is to say There is extensive debate with regard to the second generations 1,224,755 students. This article considers all the 176,160 school stu- and their composition. Ramakrishnan (2004) recommends making a dents in the city of Barcelona (Table 1), 158,056 of whom reside in distinction between children of immigrant origin born in the host the city and 14,105 in another municipality, while the place of resi- country, among the second generations (with two immigrant parents) dence of 3.999 is unknown.3 and the 2.5 generation (with one autochthonous parent) because the These have been linked with the Continuous Population Register two groups differ considerably in their sociodemographic characteris- of 1 January 2016,4 in an operation carried out by the Statistical Insti- tics. Rumbaut (2004) goes still further in adding still more complexity tute of Catalonia (Idescat), which makes it possible to recover infor- to the classification, breaking down the immigrants into several cate- mation about the students' year of arrival in Catalonia as well as their gories in keeping with the premise that generational cohorts, defined parents' place of origin and nationality. Until now, inadequate T A B L E 1 Students according to nationality, country of birth and immigration status in Barcelona, by educational stage and type of school, 2015–2016 By nationality Public school Private school All students Foreigners Total (%) Foreigners Total (%) Foreigners Total (%) Preschool (second cycle) 4,314 18,011 24.0 2,032 21,895 9.3 6,346 39,906 15.9 Primary 6,109 34,915 17.5 2,796 46,952 6.0 8,905 81,867 10.9 Secondary 4,142 19,592 21.1 2,132 34,795 6.1 6,274 54,387 11.5 Total 14,565 72,518 20.1 6,960 103,642 6.7 21,525 176,160 12.2 By country of birth Public school Private school All students Immigrant Total (%) Immigrant Total (%) Immigrant Total (%) Preschool (second cycle) 1,061 18,011 5.9 861 21,895 3.9 1,922 39,906 4.8 Primary 4,169 34,915 11.9 2,457 46,952 5.2 6,626 81,867 8.1 Secondary 5,508 19,592 28.1 3,549 34,795 10.2 9,057 54,387 16.7 Total 10,738 72,518 14.8 6,867 103,642 6.6 17,605 176,160 10.0 By migratory status (with relation to migration) Public school Private school All students Non- Non- Non- autochthonous Total (%) autochthonous Total (%) autochthonous Total (%) Preschool 8,448 18,011 46.9 5,595 21,895 25.6 14,043 39,906 35.2 (second cycle) Primary 14,113 34,915 40.4 9,743 46,952 20.8 23,856 81,867 29.1 Secondary 8,507 19,592 43.4 7,206 34,795 20.7 15,713 54,387 28.9 Total 31,068 72,518 42.8 22,544 103,642 21.8 53,612 176,160 30.4 Note: Source: Department of Education and Idescat. BAYONA-I-CARRASCO AND DOMINGO 5 of 14 availability of statistical data on student distribution has not permitted stages, with a greater number in the preschool phase when proximity this kind of analysis in Spain. The link between data from administra- to the residence is a very important factor in choice of schools. At the tive records (collected by the Department of Education) and that of same time, there is an unequal presence of public and private systems, the population register (from Idescat), on the basis of which a new sta- with larger numbers of private schools in secondary education. tistical archive has been formed, has always been made bearing in In the analysis of segregation, the dimension of evenness mind the need to safeguard their private nature before access is given (Massey & Denton, 1988) has been used. This is done on the basis of in order to respect confidentiality in their use. three indicators. The first is the segregation index (S),5 which com- Moreover, rapid acquisition of Spanish nationality by some pares the distribution of a population group with the total and is indi- groups of students selectively conceals the children of immigrants of cated in cases like the present study in which there is a high diversity some origins, which means that information by nationality is biased of origins. Its formulation is as follows: and only representative for some groups. The perspective of genera- n   tional cohort applied in this study is only possible when school records 1X  xi − ti − xi  S= are related with population records or, in the case of Spain, the Offi- 2 i = 1 X T −X  cial Population Figures. Hence, along the lines of Rumbaut's work (Rumbaut, 2004) a generational typology is constructed, combining In a second instance, the dissimilarity index (D) is used to compare the student's place of birth, age of arriving in Catalonia and birthplace population subgroups, and the distribution in space of the two groups of parents. The classification used is as follows: considered (Duncan & Duncan, 1955). Its formulation is as follows6: n   1. First generation: schoolchildren born abroad and who arrived in 1X xi yi  D=  −  Catalonia aged seven and older and who have therefore not been 2 i=1 X Y in the educational system from the start of their schooling. 2. Generation 1.75: schoolchildren born abroad but who arrived in Third, the existence of a dual education system (public and private) Catalonia before the age of seven and who therefore entered the requires consideration of a new indicator that allows quantification of the obligatory educational system at the beginning. contribution of these subsystems to total segregation. In an approach sim- 3. Second generation: schoolchildren born in Spain of two parents ilar to that of Jenkins et al. (2008), Vázquez (2012), Murillo (2016), Murillo (or one, where information about only one parent is available) who and Martínez-Garrido (2018) and Ferrando et al. (2020), we calculate the were born abroad. square root index or Hutchens indicator (Hutchens, 2004), which has the 4. Generation 2.5: schoolchildren born in Spain, one of whose parents property of additive decomposition and allows breakdown into subsys- was born abroad and the other in Spain. tems or, in this case, the influence of public and private schools on segre- 5. Autochthonous: schoolchildren born in Spain of parents also born in gation. The indicator is defined as follows: Spain. When there is information about only one of the parents and Xk   rffiffiffiffiffiffiffiffiffiffiffiffi that parent was born in Spain, the student comes under this heading. x1i x2i x1i H= − : i=1 X 1 X 2 X1 In the case of Barcelona, complete information for all students is not available as this is lacking for a small percentage (3.4%). Among them, Here, x1i and x2i represent the numbers of students in the minority the category of Other immigrants has been recovered for those born and majority groups in the school i, and X1 and X2 are the totals for abroad (1% of students) and whose year of arrival in Catalonia is not these subgroups in the municipality. This is decomposed into known. Hence, a small percentage of the total number of students is lost (2.4%, among whom 0.4% are of foreign nationality and would be H = Hwithin + Hbetween added to the second generation). It is possible that these are more complex cases, which could include a more recent arrival or a more where vulnerable situation. In Spain, school attendance is obligatory from 6 to 16 years of X G Hwithin = W g Hg age. It is divided into two stages: primary education (6 years) and sec- g=1 ondary education (ESO, 4 years). Before this, there are cycles of pre- school education, in which the second cycle (children of 3–5 years) is practically universal. After completing ESO, students can move on to and the baccalaureate (as a preliminary step to university), or to vocational sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi   ffi training courses, which are always noncompulsory. In the present Pg Rg Wg = : study, data pertaining to obligatory schooling are analysed, although P R in some of the first descriptive graphs, children from the second cycle of pre-school education are included. This differentiated analysis is Here, g refers to the subgroups, Wg to the influence of the subgroup based on the existence of different numbers of schools at different g, Pg and Prg to the number of cases (students) in the subgroup g with 6 of 14 BAYONA-I-CARRASCO AND DOMINGO regard to the majority groups P and R (public schools and private in 1996 and 8.7% in 2001 and a volume of 58,385 and 110,129 schools respectively). Although it is one of the few indices that allows immigrants, respectively, the figures have risen to a maximum of breakdown, it has the problem of showing generally low values and 25.7% and 420,955 residents in 2019. The effects of the economic also of being little known among researchers (Allen & Vignoles, 2007), crisis are barely noticeable in this evolution, except for a one-off which makes comparison and interpretation difficult. decrease in 2014, by contrast with other Spanish cities. At present, These indicators will be calculated, first, according to nationality half the immigrant population is from the Americas (50.2%), a quar- and place of origin of the students, identified by continent. The unit ter from Europe (24.8%), while the remainder is divided between of analysis employed is the primary or secondary school. In the city of Asians (18.2%) and Africans (6.7%). This is a young population that Barcelona, these schools exist in different numbers, depending on the has arrived relatively recently, coming from a wide range of places, stage of education. There are 333 primary schools, equally divided among which 42 countries each account for more than a thousand 7 between 166 public and 167 private. At secondary school level, the residents in the city. In 2018, a third of the immigrant population presence of private schools increases, and greatly so, when by con- has Spanish nationality, especially those from the Americas (48%), trast with only 65 public schools, they account for 147 (69.3%) out of while Asians (15.7%) and Europeans (13.8%) have less access, in the a total of 212 schools with a representation that is well above the latter case because, being from an EU country, they have the same average for Catalonia (47.9%). rights as Spaniards and do not need to obtain citizenship. Because Residential segregation is also calculated for the population aged this is a city that acts as a port of first entry and subsequent metro- between 6 and 15 residing in Barcelona. The same ages are therefore politan redistribution, the proportion of minors is lower than in compared, an important factor in segregation analysis (Sabater & other suburban municipalities. Catney, 2019). The calculation is made by nationality and country of As for internal distribution, half the population in the old city cen- birth with data from the Official Population Figures on 1 January tre is composed of immigrants (59.7% in El Raval and 61.0% in the 2016 and, therefore, on the same date as the school data. Two scales Gothic Quarter), with significant percentages in some of the periph- of analysis are used for this, the census section (1,068 census sec- eral lower-income neighbourhoods, for example, Ciutat Meridiana tions) and Basic Statistical Areas (BSA, with 238 areas), with averages (41.6%). In most neighbourhoods, the values are around 20%, with a of between 1,500 and 6,700 inhabitants, respectively. The latter scale minimum of 9.1% in Canyelles, a neighbourhood constructed in 1974, presents a figure for units that is closer to that for schools and hence where there is still only a small presence of international immigrants. more comparable. 4.2 | School-age population according to the 4 | S E G RE G A T I O N I N T H E S C H O O L generational perspective 4.1 | Evolution of the immigrant and foreign In the school year of 2015–2016, there were 176,160 young people population in Barcelona studying in Barcelona, 12.2% foreigners and 10.0% immigrants, or 21,525 and 17,605 respectively. These percentages are much lower Since the end of the 20th century, Barcelona has experienced a than those presenting the total of foreigners and immigrants in the continuous increase in immigration, in such a way that, from 3.9% city as a whole on the same date (16.6% and 22.5%, respectively), F I G U R E 1 Students according to nationality, place of birth and school year, 2015–2016. Source: Department of Education and Idescat BAYONA-I-CARRASCO AND DOMINGO 7 of 14 partly because of an age structure in which young adults predominate of three foreign students attend public schools. The concentration is but also owing to laws on the acquisition of Spanish nationality and less depending on the place of birth (14.8% and 6.6%, respectively), the recent evolution of migratory flows in Spain which, after a maxi- except in secondary education where 28.1% of the students in public mum number of arrivals in 2007, experienced a period when these schools are immigrants by comparison with 10.2% in private schools. figures fell, even to the point of reaching a negative migratory balance The paradox hidden behind these figures can be interpreted through in the years of greatest intensity of the economic crisis. two processes, which will later be observed in more detail: the segre- A significant part of the students of foreign nationality were born gation of some origins, independently of place of birth, in the first in Spain (47%), especially in the early years of schooling when they case, and the concentration of newly arrived students in just a few exceed 75%, while, in secondary education, the situation is precisely secondary schools, in the second. the opposite, with more than 90% born abroad (Figure 1, left). From The number of students related with the migratory process rises this standpoint, the composition of the students is dominated by to 53,612 or 30.4% of all students when the immigration status typol- Asians (36.5%), with 28.5% from the Americas, 25.5% from Europe, ogy is used (Table 2). By educational stages, the numbers are 14,043 and 9.5% from Africa. Among the students of immigrant origin, how- for pre-school (35.2%), 23,856 for primary (29.1%) and 15,713 ever, access to nationality is significant (35.4%), with smaller varia- (28.9%) for secondary. By type of school, the gap between systems tions between the different levels of education (from a minimum of continues to grow, with 42.8% of the students in public schools and 26.8% to a maximum of 43.5%) and with a profile in which students 21.7% in private schools. These percentages are the result of a com- born in the Americas predominate (44.8%), followed by 28.4% born in position where not all the categories have the same weight. With the Asia, 20.4% in Europe, and 6.3% in Africa. private school, for example, Generation 2.5 (for which one of the par- When a distinction is made in accordance with the type of school ents is autochthonous) represents one third of the students of immi- (public or private), there is a clear overrepresentation of immigrant grant origin, when in public schools the figure is less than 17%. In this students in public schools (Table 1), and this is particularly the case regard, it is notable that 69.3% of the first generations attend public with foreign students (20.1% of students in public schools and 6.7% schools by comparison with 33% of autochthonous students. These of those in private schools). This is the result of the fact that two out differences are clearly observed in Figure 2 where, on an educational TABLE 2 Index of dissimilarity according to migratory status and educational stage, Barcelona year 2015–2016 Secondary education Autochthonous First generation Generation 1.75 Second generation Generation 2.5 Primary education Autochthonous - 0.53 0.33 0.41 0.20 First generation 0.56 - 0.29 0.30 0.54 Generation 1.75 0.42 0.26 - 0.20 0.34 Second generation 0.48 0.28 0.23 - 0.42 Generation 2.5 0.19 0.52 0.38 0.44 - Note: Source: Education Department and Idescat. F I G U R E 2 Proportion of students according to immigration status in Barcelona by school year and type of school, 2015–2016. Source: Department of Education and Idescat 8 of 14 BAYONA-I-CARRASCO AND DOMINGO level by educational level basis and in accordance with ownership of segregation. Nevertheless, some phenomena diverge, for example, the school, the changing composition according to the migrant status increased school segregation in the last year of secondary education, of students is represented. which is observed in some cases at school, although it is not reflected in territorial terms. 4.3 | School and residential segregation in Barcelona 4.4 | Segregation according to immigration status The levels of school and residential segregation are in the middle Using immigration status offers a new reading of segregation that range. Using the census section, residential segregation is very stable complements the partial views provided by data by origin and nation- with age and is estimated as being above 0.4. Depending on place of ality. Hence, segregation shows very different intensities according to birth, segregation decreases with age, especially in the early ages. the generational typology (Figure 5). For first-generation students, These values drop to below 0.3 if BSA are used. The relationship segregation reaches its highest levels, even as much as twice those between segregation by nationality or place of birth is maintained. At observed for the 2.5 generation. These values are situated at around school, segregation by nationality is generally greater and practically 0.5 (in the first school years there are few students), and they rise sig- stable through all educational levels. However, analysis by country of nificantly in the last school year. There is a considerable distance from birth draws attention to the increase observed in the final years of Generation 1.75, indicating both greater territorial concentration of secondary education when, unlike what is happening from the resi- students arriving in Barcelona and also bad management of year- dential standpoint, segregation considerably increases. In the compari- round enrolment, which is to say allocation to schools of students son between residence and school, segregation is greater at school, a who frequently arrive when the school year has already started. It is situation that is particularly visible if the geographic unit of compari- important to note, too, how the second generation experiences levels son is the BSA (Figure 3). Because BSA has a number of units that is of segregation that are even higher than those for students of Gener- more similar to that for schools, the values are more comparable and ation 1.75. This is a matter of concern when 60% of these students allow us to state that levels of segregation in schools are higher than have Spanish nationality and when socioeconomic differences as well those for residence. as statistical invisibility can partly explain these values. Then again, The same comparison is carried out among continental groups, there is hardly any segregation for Generation 2.5, with values that with some interesting results (Figure 4). First, comparing the form of are always less than 0.3. There is a big distance between them and the graphs and their evolution, one can say that there is a clear corre- the second generation, which does not vary between years of school- lation between school and residential segregation, as the dynamics ing, and hence the need to distinguish between the two subgroups of between ages and educational levels are similar. Comparing school students. and residence, and focussing the analysis on BSA data, one again sees As for distribution between groups in schools, the dissimilarity greater school segregation. In line with the previous literature, the index enables comparison (Table 2, rows). In primary schools, autoch- greatest differences are observed among immigrants from the Ame- thonous groups and Generation 2.5 are the closest groups, and there ricas. If their values for residence are lower, their segregation in are also similar distributions among Generation 1.75 and the second schools is greater than it is among Europeans. generation. By contrast, the greatest difference is between the first- Beyond the classical problem of the number of units, it is generation and autochthonous students and also Generation 2.5. Also observed that higher values for Africans, or the diminishing segrega- notable is the difference between the autochthonous group and the tion of students from the Americas, are similar between residence and second generation, which is even greater than it is with Generation school. Hence, school segregation is partly explained by residential 1.75. In secondary education (Table 2, columns), most relationships F I G U R E 3 School segregation by nationality and origin, Barcelona, 2015–2016. Source: Departament d'Ensenyament and Idescat BAYONA-I-CARRASCO AND DOMINGO 9 of 14 F I G U R E 4 School and residential segregation by continental origin and nationality. School year 2015–2016. Source: Department of Education and Idescat indicate a drop in segregation, except for the first generations where there is mostly an upward trend. 4.5 | Segregation and public and private schools From the generational standpoint, the segregation values differ between public and private schools. The segregation index, calculated by stage of education, for schools as a whole, and by ownership (public or private), indicates how, in the cases of the first and second generations, segregation is greatest in private schools, coinciding with the highest values of segregation (Figure 6). In the remainder of the cases, segregation is greater in public schools. This is due to an F I G U R E 5 School segregation according to immigration status and school year 2015–2016. Source: Education Department and unequal distribution of students between public and private schools Idescat and because the presence of students of immigrant origin in private 10 of 14 BAYONA-I-CARRASCO AND DOMINGO F I G U R E 6 School segregation by migratory status, school stage and type of school, 2015–2016. Source: Education Department and Idescat schools is much more polarised than it is in public schools. This state-subsidised private schools, which increases segregation within happens, in particular, in some religious state-subsidised private the same system (in this case, private). In addition, and owing to schools where the majority of the students are of immigrant origin. In different strategies for choosing centres, it is notable how some cases, for example, those of Moroccans and Pakistanis, autochthonous students experience greater levels of segregation vis- about 90% of the students go to public schools while, in others, like à-vis the students as a whole than the 1.75 and 2.5 generations. Filipinos, the figure is at the same level as that for Spanish students Finally, the Hutchens' index allows evaluation of each one of the with regard to private schools. Moreover, some of these groups subsystems with regard to total segregation, which is the third aim of are to be found in just a few schools, many of them religious this study. The values within which this indicator moves are TABLE 3 School segregation (Hutchens index) by students according to educational stage, Barcelona Primary school H intra-subsystems Public Private H total H gross Weight Contribution H gross Weight Contribution H intersystems Nationality 0.1576 0.0891 0.5204 0.0464 0.1550 0.4359 0.0676 0.0437 Country of birth 0.0985 0.0700 0.5071 0.0355 0.0819 0.4682 0.0384 0.0246 First generation 0.1866 0.1103 0.5423 0.0598 0.2080 0.4179 0.0869 0.0398 Generation 1.75 0.0796 0.0654 0.5044 0.0330 0.0577 0.4765 0.0275 0.0191 Second generation 0.1337 0.0745 0.5036 0.0375 0.1333 0.4618 0.0616 0.0346 Generation 2.5 0.0393 0.0654 0.4338 0.0284 0.0369 0.5661 0.0209 0.0002 Secondary school H intra-subsystems Public Private H total H gross Weight Contribution H gross Weight Contribution H intersystems Nationality 0.1714 0.0884 0.4604 0.0407 0.1487 0.4803 0.0714 0.0593 Country of birth 0.1165 0.0789 0.4348 0.0343 0.0704 0.5197 0.0366 0.0456 First generation 0.1819 0.1016 0.4783 0.0486 0.1451 0.4544 0.0659 0.0673 Generation 1.75 0.0482 0.0233 0.4270 0.0100 0.0393 0.5567 0.0219 0.0164 Second generation 0.0957 0.0297 0.4339 0.0129 0.1150 0.5462 0.0628 0.0199 Generation 2.5 0.0413 0.0465 0.3424 0.0159 0.0372 0.6567 0.0244 0.0009 Note: Source: Department of Education and Idescat. BAYONA-I-CARRASCO AND DOMINGO 11 of 14 significantly smaller than those provided by the segregation indicator and Asians correlates with high levels of residential segregation for and, although they are not directly comparable, on this occasion the these origins, and, on the other hand, the evolution of segregation in interest is to find out how segregation is divided between the school school years is similar to that observed with ages in the territory. With systems, and the weight of each one in total segregation (Table 3). some particularities, school segregation depends on residential segre- The segregation values repeat some previously found consisten- gation, which means that part of it is explained by the distribution of cies, for example, greater segregation of the first and second genera- the immigrant population around the city. One of the exceptions tions. Generation 2.5, especially in primary school, shows the lowest noted is what happens in the last year at school when there is an values. However, the differences between calculation by nationality increase of segregation that is not visible in territorial terms. This is a and by country of birth are magnified, with much greater segregation consequence of bad management of ‘live enrolment’ that results in a in the former case. concentration of students in a few secondary schools and especially Results concerning the contribution to segregation of the dual of those who arrive in the city as part of the family reunification pro- school system indicate how segregation of students in state-funded gramme with the intention of swift incorporation into the labour mar- private schools is greater than in public schools for most of the groups ket. These students belong to the first generation and their analysed, and the greater the segregation, the more striking this is. characteristics are very different from those of other students of the As for contribution to total segregation, in general, this is more pre- same nationality. The results, here, underscore the need to consider sent in private than in public schools, especially when it comes to higher aspects pertaining to the migratory process when analysing school levels of segregation. A third element influencing segregation is differ- segregation—as we have done in the present study—and the impor- ent distribution between two systems. This component represents tance of giving attention to certain specific groups. about a quarter of segregation in primary schools, while, in secondary One of the limitations of this study is that we have not had access education, it accounts for more than a third in most cases. In primary to socioeconomic data on students and their families, which would education, and for second-generation students, the result of the differ- have enabled us to carry out analysis by social class. We are aware of ent distribution of students between systems gives a figure of 25.9%. the importance of social class in segregation levels (see Prieto-Latorre In secondary education, and for first-generation students, this value et al., 2020, for the Spanish region of Andalusia), but the aim of our rises to 37%. It is only in the 2.5 generation that the differences study is to draw attention to the impact of cohort and origin in segre- between systems would contribute nothing to total segregation. gation levels. It is also an example of the importance for social research of being able to use increasingly detailed information like that which can be derived from ever more common practices of taking 5 | C O N CL U S I O N S advantage of administrative records. By and large, the generational perspective employed in the present The usual analyses, based on nationality, of residential segregation of study indicates how students of the first generation experience the students of immigrant origin present biased views of segregation, as highest levels of segregation. And these levels of segregation, which the presence of these students in the statistics is the result of casu- are higher than those found in relation with nationality, are in large part istry regarding legislation on acquisition of nationality. Hence, in the explained by the concentration that presently exists in the public- city of Barcelona, students who are descendants of immigrants of school system and, within that, in a few secondary schools. Manage- Latin American origin, or who are themselves immigrants from Latin ment of student enrolment in secondary schools is not only deficient, America, are not counted in the observation, while those of other, for but it gives rise to high levels of segregation. Hence, adequate policies example, Asian or African, origins remain in the statistics. Moreover, for receiving and distributing students who join classes during the there is also a selection that makes visible those who have arrived school year could be very helpful in attenuating these high levels. more recently, while others, who have been in the country for years, The second group experiencing high levels of segregation is that disappear from view. The present study's cohort perspective derived of the second generation, which suggests perpetuation of socioeco- from links between statistical records multiplies the number of nomic differentiation from parents to children. Their statistical students under consideration and allows analysis in four different visibilisation is just a first step towards policies concerning this group, categories. This, then, offers a more complete account of the relation- especially in the city's more vulnerable areas. This greater segregation ship between the migratory process and school segregation and could is in keeping with the studies by Schindler (2007) and contradicts the be important for the implementation of public policy in this regard. view deriving from assimilationist ideas of greater dispersion among Analyses comparing school and residential segregation indicate, descendants. Only children of mixed couples (Generation 2.5) experi- first of all, the relationship between the two, this being derived from ence lower levels of segregation. However, the segregation of Gener- observation of the temporal evolution by level of education and ages ation 1.75 appears at lower levels than those observed for the second of school and residential segregation. As in other contexts (Johnston generation. This could be due to their greater statistical visibility et al., 2004; Schindler, 2007), school segregation is generally of greater (by comparison with the second generation, they can be identified by intensity. This is observed when comparing the patterns by nationality place of birth), which means that they receive government attention or place of birth with very similar dynamics between school and resi- and some focus in policies. Otherwise, this higher visibility is the result dence. On the one hand, considerable school segregation of Africans of their residential concentration in neighbourhoods that have been 12 of 14 BAYONA-I-CARRASCO AND DOMINGO given more attention because of their considerable percentages of RE FE RE NCE S foreign-born populations. The results could also be due to a composi- Alegre, M. A. (2017). Polítiques de tria i assignació d'escola: Quins efectes tion effect in that the composition by origin of the different categories tenen sobre la segregació escolar?. Barcelona: Fundació Jaume Bofill. Alegre, M. A., Benito, R., & González, I. (2008). Procesos de segregación y is influenced by immigration flows that change over time. As for our polarización escolar: la incidencia de las políticas de zonificación esco- initial hypothesis on the existence of gradation between segregation lar. Profesorado. Revista de currículum y formación del profesorado, and generations, this is partially confirmed, with the exception of 12(2). Generation 1.75, as we have noted. Detailed analysis of national Allen, J. P., & Turner, E. (1996). Spatial Patterns of Immigrant Assimilation. The Professional Geographer, 48(2), 140–155. https://doi.org/10.1111/ origin and generation deserves attention in future research. j.0033-0124.1996.00140.x Finally, the analysis carried out with regard to the ownership status Allen, R., & Vignoles, A. (2007). What should an index of school segrega- of schools sheds light on the effect on segregation of the dual school tion measure? Oxford Review of Education, 33(5), 643–668. https://doi. system. The results indicate heterogeneity in the origin of segregation. org/10.1080/03054980701366306 Aparicio, R., & Portes, A. (2014). Crecer en España. La integración de los Hence, a third of presently existing segregation is explained by differ- hijos de inmigrantes. Barcelona: La Caixa. Estudios Sociales, 38. ences generated by the dual school system and even more notably in Arango, J. (2009). Después del gran boom: La inmigración en la bisagra del secondary education, which has a smaller number of schools. At the cambio. In E. Aja, J. Arango, & J. Oliver (Eds.), La inmigración en tiempos same time, the results suggest that internal differences within a same de crisis. Anuario de la inmigración en España (pp. 52–73). Barcelona: system are still important, and these should also be a primary focus of Fundació CIDOB. Arbaci, S. (2019). Paradoxes of segregation: Housing systems, welfare regimes attention in both public and state-funded private schools. and ethnic residential change in southern European cities. Hoboken: John Wiley & Sons. https://doi.org/10.1002/9781118867365 ACKNOWLEDGEMEN TS Arbaci, S., & Malheiros, J. (2010). De-segregation, peripheralisation and This paper is part of the Project ‘Demographics, Migrations and New the social exclusion of immigrants: Southern European cities in the 1990s. Journal of Ethnic and Migration Studies, 36(2), 227–255. https:// Statistical Frontiers: Big Data, Continuous Population Registers, and doi.org/10.1080/13691830903387378 Administrative Records’ (CSO2017-85670-R), funded by the Ministry Bayona, J., & Domingo, A. (2018). School failure among descendants of of Economy and Competitiveness. immigrants in Catalonia: More than just another assignment. Perspec- tives Demogràfiques, 11, 1–4. Bayona, J., & Domingo, A. (2019). Proceso migratorio, concentración esco- CONF LICT OF IN TE RE ST lar y resultados académicos en Cataluña. Migraciones, 47, 3–34. There is no conflict of interest. https://doi.org/10.14422/mig.i47y2019.001 Bayona, J., & López, A. (2011). Concentración, segregación y movilidad ORCID residencial de los extranjeros en Barcelona. Documents d'Anàlisi Geo- gràfica, 57(3), 381–412. https://doi.org/10.5565/rev/dag.234 Jordi Bayona-i-Carrasco https://orcid.org/0000-0003-2819-9085 Benito, R., & Gonzàlez, I. (2007). Processos de segregació escolar a Catalu- Andreu Domingo https://orcid.org/0000-0003-3270-1939 nya. Barcelona: Mediterrània. Bonal, X. (2012). Education policy and school segregation of migrant stu- ENDNOTES dents in Catalonia: The politics of non-decision-making. Journal of Edu- 1 cation Policy, 27(3), 401–421. https://doi.org/10.1080/02680939. The Síndic de Greuges is the Ombudsman of Catalonia. 2011.645168 2 With regard to the two kinds of private schools, the available database Bonal, X., & Bellei, C. (2018). The renaissance of school segregation in a does not let us distinguish, between those that are state subsidised and context of globalization. In X. Bonal & C. Bellei (Eds.), Understanding those that are not. school segregation: Patterns, causes and consequences of spatial inequal- 3 The data cover is good. The differences with data published by the ities in education. London: Bloomsbury. Barcelona City Council for the same school year are of the order of 0.1%. Bonal, X., & Zancajo, A. (2020). Elección de Escuela, Movilidad y 4 This operation, based on the Official Population Figures, showing all resi- Segregación Escolar del Alumnado Vulnerable en Barcelona. REICE. dents of a municipality independently of the legal situation, represents Revista Iberoamericana Sobre Calidad, Eficacia Y Cambio en Educación, data crossing from administrative records carried out by the Institut 18(4), 197–218. https://doi.org/10.15366/reice2020.18.4.008 d'Estadística de Catalunya (Idescat—Statistical Institute of Catalonia). Bonal, X., Zancajo, A., & Scandurra, R. (2019). Residential segregation and 5 school segregation of foreign students in Barcelona. Urban Studies, 56 Where xi is the population of group X in the spatial unit i; X, the popula- (15), 3251–3273. https://doi.org/10.1177/0042098019863662 tion of group X in the municipality; ti, the total population in spatial unit Bonal, X., Zancajo, A., & Scandurra, R. (2020). Student mobility and school i; T, the total population of the municipality, where n is the number of segregation in an (un)controlled choice system: A counterfactual spatial units. approach. British Educational Research Journal. https://doi.org/10. 6 Where xi is the population of group X in the spatial unit i; X, the popula- 1002/berj.3694 tion of group X in the municipality; yi, the population of the group Y in Boterman, W. (2019). The role of geography in school segregation in the the spatial unit i; Y, the population of group Y in the municipality; and n, free parental choice context of Dutch cities. Urban Studies, 56(15), the number of spatial units. 3074–3094. https://doi.org/10.1177/0042098019832201 7 Most of the latter are government-subsidised centres, which is to say Boterman, W., & Musterd, S. (2016). Cocooning urban life: Exposure to they are privately managed but partially funded by the Generalitat diversity in Neighbourhoods, workplaces and transport. Cities, 59, (Government) of Catalonia. In the Barcelona City Council data, only eight 139–147. https://doi.org/10.1016/j.cities.2015.10.018 private primary schools and seven secondary schools are not state Boterman, W., Musterd, S., Pacchi, C., & Ranci, C. (2019). School segrega- subsidised. tion in contemporary cities: Socio-spatial dynamics, institutional BAYONA-I-CARRASCO AND DOMINGO 13 of 14 context and urban outcomes. Urban Studies, 56(15), 3055–3073. López-Falcón, D., & Bayona, J. (2012). Segregación escolar y residencial en https://doi.org/10.1177/0042098019868377 Barcelona: del boom migratorio al asentamiento. In F. J. García-Cast- Calero, J., & Escardíbul, J. O. (2016). Proceso educativo y resultados del año & A. Olmos (Eds.), Segregaciones y construcción de la diferencia en alumnado nativo y de origen inmigrante en España. Un análisis basado la escuela (pp. 21–42). Editorial Trotta. en PISA-2012. Estudios de economía aplicada, 34(2), 413–438. Marcinczak, S., Tammaru, T., Strömgren, M., & Lindgren, U. (2015). Chang- Carrasco, S., Pàmies, J., Ponferrada, M., Ballestín, B. & Bertran, M. (2007). ing patterns of residential and workplace segregation in the Stockholm Segregación escolar e inmigración en Cataluña: aproximacions metropolitan area. Urban Geography, 36(7), 969–992. https://doi.org/ etnográficas. EMIGRA Working Paper, 126, 1–26. 10.1080/02723638.2015.1012364 Cebolla-Boado, H., & Garrido, L. (2011). The impact of immigrant concen- Martínez Del Olmo, A., & Leal, J. (2008). La segregación residencial, un tration in spanish schools: school, class, and composition effects. indicador espacial confuso en la representación de la problemática European Sociological Review, 27(5), 606–623. https://doi.org/10. residencial de los inmigrantes económicos. Architecture, City and Envi- 1093/esr/jcq024 ronment, 8, 53–64. De Miguel, V., & Solana, M. (2017). Immigrants in the educational system Martori, J. C., Hoberg, K., & Surinach, J. (2006). Población inmigrante y in Spain: Who persists? Social Indicators Research, 132, 733–755. espacio urbano: Indicadores de segregación y pautas de localización. Domingo, A., & Bayona, J. (2021). Trajectòries migratòries i rendiment EURE, 32(97), 49–62. escolar als barris de Barcelona. Documents d'Anàlisi Geogràfica, 67(1), Massey, D., & Denton, N. (1988). The dimensions of residential segrega- 73–102. https://doi.org/10.5565/rev/dag.608 tion. Social Forces, 67(2), 281–315. https://doi.org/10.2307/2579183 Domingo, A., & Cabré, A. (2015). La demografía del siglo XXI: Evolución Murillo, F. J. (2016). Midiendo la Segregación escolar en América Latina. reciente y elementos prospectivos. In C. Torres (Ed.), España 2015. Un análisis metodológico utilizando el TERCE. Revista Iberoamericana Situación social (pp. 63–73). Madrid: Centro de Investigaciones Sobre Calidad, Eficacia Y Cambio en Educación, 14(4), 33–60. Sociológicas. Murillo, F. J., Belavi, G., & Pinilla, L. M. (2018). Segregación escolar Duncan, O., & Duncan, B. (1955). A methodological analysis of segregation público-privada en España. Papers. Revista de Sociologia, 103(3), indices. American Sociological Review, 20(2), 210–217. https://doi.org/ 307–337. https://doi.org/10.5565/rev/papers.2392 10.2307/2088328 Murillo, F. J., & Martínez-Garrido, C. (2018). Impact of the economic crisis Ferrando, F., Hernández-Almeida, M., Oreiro, C., Seijas, M. N., & on school segregation in Spain. Revista de Educación, 381, 61–87. Urraburu, J. (2020). Evolución de la Segregación Socioeconómica en la Musterd, S., Marcin czak, S., Van Ham, M., & Tammaru, T. (2017). Socioeco- Educación Pública de Uruguay. REICE. Revista Iberoamericana Sobre nomic segregation in European capital cities: Increasing separation Calidad, Eficacia Y Cambio en Educación, 18(4), 143–169. https://doi. between poor and rich. Urban Geography, 38(7), 1062–1083. https:// org/10.15366/reice2020.18.4.006 doi.org/10.1080/02723638.2016.1228371 Ferrer, F., Castejón, A., Castel, J. L., & Zancajo, A. (2011). PISA 2009: Nordin, M. (2013). Immigrant school segregation in Sweden. Population Avaluació de les desigualtats educatives a Catalunya. Barcelona: Fun- Research and Policy Review, 32, 415–435. https://doi.org/10.1007/ dació Jaume Bofill. s11113-013-9271-z Fiel, J. E., & Zhang, Y. (2018). Three dimensions of change in school segre- Oberti, M., & Savina, Y. (2019). Urban and school segregation in Paris: The gation: A grade-period-cohort analysis. Demography, 55, 33–58. complexity of contextual effects on school achievement: The case of Galeano, J., & Bayona, J. (2018). Residential segregation and clustering middle schools in the Paris metropolitan area. Urban Studies, 56(15), dynamics of migrants in the metropolitan area of Barcelona: A demo- 3117–3142. https://doi.org/10.1177/0042098018811733 spatial analysis at the census tract level. Revue Quetelet, 6(1), 99–127. Portes, A., Aparicio, R., & Haller, W. (2016). Spanish legacies: The coming of https://doi.org/10.14428/rqj2018.06.01.05 age of the second generation. Oakland, California: University of California García-Castaño, F. J., & Carrasco, S. (2011). Población inmigrante y Press. https://doi.org/10.1525/california/9780520286290.001.0001 escuela: conocimientos y saberes de investigación. Ministerio de Prieto-Latorre, C., Marcenaro-Gutierrez, O., & Vignoles, A. (2020). School educación. Colección Estudios Creade, 8, 269–288. segregation in public and semiprivate primary schools in Andalusia. González-Ferrer, A., & Cebolla-Boado, H. (2018). Los hijos de la British Journal of Educational Studies, 1–22. https://doi.org/10.1080/ inmigración en España: valores, aspiraciones y resultados. In A. Blanco 00071005.2020.1795078 (Ed.), Informe España 2018 (pp. 111–163). Madrid: Universidad Pontifi- Ramakrishnan, S. K. (2004). Second-generation immigrants? The ‘2.5 gen- cia Comillas. eration’ in the United States. Social Science Quarterly, 85(2), 380–399. Gramberg, P. (1998). School segregation: The case of Amsterdam. Urban https://doi.org/10.1111/j.0038-4941.2004.08502013.x Studies, 35(3), 547–564. https://doi.org/10.1080/0042098984907 Rathelot, R., & Safi, M. (2014). Local ethnic composition and natives' and Hutchens, R. (2004). One measure of segregation. International Economic Review, Immigrants' geographic mobility in France, 1982–1999. American 45(2), 555–578. https://doi.org/10.1111/j.1468-2354.2004.00136.x Sociological Review, 79(1), 43–64. https://doi.org/10.1177/ Jenkins, S. P., Micklewright, J., & Schnepf, S. V. (2008). Social segregation 0003122413514750 in secondary schools: How does England compare with other coun- Robinson, V. (1984). Asians in Britain: A study in encapsulation and mar- tries? Oxford Review of Education, 34(1), 21–37. https://doi.org/10. ginality. In C. Clarke, D. Ley, & C. Peach (Eds.), Geography & ethnic plu- 1080/03054980701542039 ralism (pp. 231–257). George Allen & Unwin: London. Johnston, R., Burgess, S., Harris, R., & Wilson, D. (2008). ‘Sleep-walking Rubiales, M. (2020). Segregación en las metrópolis españolas 2001–2011: towards segregation’? The changing ethnic composition of English Un análisis con detalle territorial. Documents d'Anàlisi Demogràfica, 66 schools, 1997–2003: An entry cohort analysis. Transactions of the Insti- (1), 83–105. https://doi.org/10.5565/rev/dag.581 tute of British Geographers, 33(1), 73–90. https://doi.org/10.1111/j. Rumbaut, R. (2004). Ages, life stages, and generational cohorts: 1475-5661.2007.00285.x Decomposing the immigrant first and second generations in the Johnston, R., Burgess, S., Wilson, D., & Harris, R. (2007). School and resi- United States. International Migration Review, 38(2), 1160–1205. dential ethnic segregation: An analysis of variations across England's Sabater, A., & Catney, G. (2019). Unpacking summary measures of ethnic local education authorities. Regional Studies, 40(9), 973–990. https:// residential segregation using an age group and age cohort perspective. doi.org/10.1080/00343400601047390 European Journal of Population, 35(1), 161–189. https://doi.org/10. Johnston, R., Wilson, D., & Burgess, S. (2004). School segregation in multi- 1007/s10680-018-9475-3 ethnic England. Ethnicities, 4(2), 237–265. https://doi.org/10.1177/ Sánchez-Hugalde, A. (2009). La segregació escolar dels immigrants a Cata- 1468796804042605 lunya. Quaderns d'avaluació, 13, 55–76. 14 of 14 BAYONA-I-CARRASCO AND DOMINGO Sarasa, S., Porcel, S., Navarro, L., & Thiers, J. (2018). Desindustrialización y Vázquez, E. (2012). Segregación escolar por nivel socioeconómico. Mid- crisis en la Barcelona posolímpica: ¿Hacia una ciudad dual? Eure, 44 iendo el fenómeno y explorando sus determinantes. Documento de (131), 29–52. Trabajo, 128, Universidad Nacional de La Plata. Schindler, B. (2007). Living and learning separately? Ethnic segregation of White, M. J., Biddlecom, A. E., & Guo, S. (1993). Immigration, naturaliza- school children in Copenhagen. Urban Studies, 7, 1329–1354. tion, and residential assimilation among Asian Americans in 1980. Síndic de Greuges. (2008). School segregation in Catalonia. Catalunya: Social Forces, 72(1), 93–117. https://doi.org/10.2307/2580161 Síndic de Greuges. Wilson, D., & Bridge, G. (2019). School choice and the city: Geographies of Síndic de Greuges. (2016). La segregació escolar a Catalunya (I): La gestió del allocation and segregation. Urban Studies, 56, 3198–3215. https://doi. procés d'admissió d'alumnat. Catalunya: Síndic de Greuges. org/10.1177/0042098019843481 Sorando, D., & Leal, J. (2019). Distant and unequal: The decline of social mixing in Barcelona and Madrid. REIS, Revista Española de Investigaciones Sociológicas, 167, 125–148. Tammaru, T., Marcinczak, S., Van Ham, M., Musterd S (eds) (2016). Socio- How to cite this article: Bayona-i-Carrasco J, Domingo A. economic segregation in European capital cities: East meets west. London School segregation of migrants and their descendants in a dual and New York: Routledge. school system: The case of Barcelona. Popul Space Place. Tarabini, A., Montes, A., & Parcerisa, L. (2018). Les aliances Magnet: inno- 2021;e2446. https://doi.org/10.1002/psp.2446 vació per combatre la segregació escolar (Vol. 68). Fundació Jaume Bofill, Informes Breus.