Sibling correlations and intergenerational mobility across immigrant groups

Colagrossi, Marco; Deiana, Claudio; Geraci, Andrea; Giua, Ludovica; Mazzarella, Gianluca Working Paper Sibling correlations and intergenerational mobility across immigrant groups JRC Working Papers in Economics and Finance, No. 2024/4 Provided in Cooperation with: Joint Research Centre (JRC), European Commission Suggested Citation: Colagrossi, Marco; Deiana, Claudio; Geraci, Andrea; Giua, Ludovica; Mazzarella, Gianluca (2024) : Sibling correlations and intergenerational mobility across immigrant groups, JRC Working Papers in Economics and Finance, No. 2024/4, European Commission, Ispra This Version is available at: https://hdl.handle.net/10419/311089 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. https://creativecommons.org/licenses/by/4.0/ Sibling Correlations and Intergenerational Mobility across Immigrant Groups Colagrossi, M., Deiana, C., Geraci A., Giua L., Mazzarella, G. 2024 JRC Working Papers in Economics and Finance, 2024/4 This publication is a Working Paper by the Joint Research Centre (JRC), the European Commission’s science and knowledge service. It aims to provide evidence-based scientific support to the European policymaking process. Working Papers are prepublication versions of technical papers, academic articles, book chapters, or reviews. Authors may release working papers to share ideas or to receive feedback on their work. This is done before the author submits the final version of the paper to a peer reviewed journal or conference for publication. Working papers can be cited by other peer-reviewed work. The contents of this publication do not necessarily reflect the position or opinion of the European Commission. Neither the European Commission nor any person acting on behalf of the Commission is responsible for the use that might be made of this publication. For information on the methodology and quality underlying the data used in this publication for which the source is neither Eurostat nor other Commission services, users should contact the referenced source. The designations employed and the presentation of material on the maps do not imply the expression of any opinion whatsoever on the part of the European Union concerning the legal status of any country, territory, city or area or of its authorities, or concerning the delimitation of its frontiers or boundaries. Contact information Name: Marco Colagrossi Address: Via E. Fermi 2749 – I-21027 – Ispra (Va) - ITALY Email: [email protected] Tel.: +39 0332 78 9526 EU Science Hub https://joint-research-centre.ec.europa.eu JRC138162 © European Union, 2024 The reuse policy of the European Commission documents is implemented by the Commission Decision 2011/833/EU of 12 December 2011 on the reuse of Commission documents (OJ L 330, 14.12.2011, p. 39). Unless otherwise noted, the reuse of this document is authorised under the Creative Commons Attribution 4.0 International (CC BY 4.0) licence (https://creativecommons.org/licenses/by/4.0/). This means that reuse is allowed provided appropriate credit is given and any changes are indicated. How to cite this report: European Commission, Joint Research Centre, Colagrossi, M., Deiana, C., Geraci A., Giua L., Mazzarella, G., Sibling Correlations and Intergenerational Mobility across Immigrant Groups, JRC Working Papers in Economics and Finance, 2024/4, European Commission, Ispra, Italy, 2024, JRC138162 Sibling Correlations and Intergenerational Mobility across Immigrant Groups Marco Colagrossi1, Claudio Deiana1,2,3, Andrea Geraci∗1,4, Ludovica Giua1,2, and Gianluca Mazzarella1,5 1European Commission, Joint Research Centre (JRC) 2University of Cagliari and CRENoS 3IZA 4University of Pavia 5University of Milan Abstract We study immigrant assimilation in terms of earnings dynamics and patterns of intergenerational transmission in permanent earnings by immigrant generation and neighborhood segregation levels. We estimate comparable sibling correlations across native and immigrant groups, but these seem to be explained by different factors. As immigrants assimilate, their intergenerational transmission mechanisms also become similar to natives. However, less assimilated immigrants experience weaker earnings transmission, a higher persistence of neighborhood effects, and a slower assimilation trajectory. Keywords: Immigrant assimilation; Sibling correlation; Intergenerational transmission; Segregation. JEL codes: J15; J61; J62. ∗Corresponding author: [email protected] Opinions expressed herein are those of the authors only and do not reflect the views of, or involve any responsibility for, the institutions to which they are affiliated. Any errors are the fault of the authors only. 1 1 Introduction Upon arrival in a host country, immigrants typically experience worse economic performance than natives. They often work in low-skill, low-income jobs, reside in socioeconomically or ethnically segregated neighborhoods, and come across learning and language difficulties. These disadvantages tend to be passed upon their offspring (Borjas,1993;Dustmann,2008;Liebig & Widmaier,2009;Algan et al.,2010). Given the substantial and growing proportion of immigrants in many countries, their contribution to the host economy, and the potential challenges they face, it is critical to understand what drives the immigrant-natives gap in socio-economic status and its persistence across generations. This is especially important when designing immigration policies that aim at fostering immigrant assimilation in the host country. The degree of association between the social status of different members of the same household has been extensively used in the literature to investigate the role of family background in the distribution of social and economic outcomes. Following the seminal contribution of Becker & Tomes (1976,1979), significant attention has been devoted to the analysis of vertical relations. This entails investigating the transmission of social status from parents to offspring, commonly referred to as intergenerational correlation (IGC). At the same time, several authors have looked at horizontal relationships and, in particular, at sibling correlations. A relevant feature of sibling correlation in income is that, either through its decomposition or estimating it jointly with parental income, it permits to estimate the importance of the intergenerational transmission of permanent income separately from other factors that siblings share but are orthogonal to parental income. Sibling correlation is often described as an omnibus measure of family background (e.g., parental resources, aspirations, and cultural inheritance) and community effects, such as school quality, sibling interactions, and neighborhood and network influences (Bj¨orklund et al.,2002;Mazumder,2011;J¨antti & Jenkins, 2015).1The idea is that if family and community background have little effect on economic outcomes, siblings will exhibit roughly the same resemblance as would a randomly drawn pair of individuals from the same birth cohort (Solon,1999). Thus, if inequalities across groups exist, measuring sibling correlation can help to identify the extent to which these inequalities are perpetuated by differences in family and community characteristics (Solon et al.,1991; Bj¨orklund et al.,2004;J¨antti & Jenkins,2015;Bj¨orklund & J¨antti,2020). We exploit this approach to study the earnings dynamics and the degree of intergenerational persistence for immigrants. Specifically, we seek to address the following questions: do immigrants exhibit similar levels of sibling correlation to natives, and is their degree of intergenerational mobility comparable? And do patterns of intergenerational transmission evolve across generations? Our results suggest that sibling correlations are similar across groups but this is 1From a policy perspective, it is key to understand (relative) social mobility (i.e., intergenerationally transmitted inequality) and the across-country differences that are observed. In this context, sibling correlations allow to estimate how much of the observed inequality can be explained by all the factors that are shared by siblings – hence beyond income. 2 due to heterogeneous patterns of intergenerational transmission across immigrant groups and across generations, which implies the existence of different assimilation dynamics. Models estimating the importance of family background typically assume that the intergenerational transmission process is homogeneous across families (e.g., Solon et al.,1991;Bj¨orklund et al.,2002,2009;Levine & Mazumder,2007;Mazumder,2008;Bj¨orklund et al.,2010). These studies usually find higher levels of sibling correlation in the United States (namely, 0.40–0.50) compared to European Nordic countries (ranging from 0.14 to 0.26), and that family background has limited weight on the sibling correlation. For instance, Bj¨orklund & J¨antti (2009) estimate that the intergenerational component accounts for only 6% of the sibling correlation in Denmark. By relaxing such assumption and formalising a heterogeneous intergenerational transmission process, Bingley & Cappellari (2019) recently showed that the role of family background may be more significant than previous studies indicate. Using Danish administrative data, they found a sibling correlation of 0.22, which originates mostly from the intergenerational transmission process (72%). The use of a more general model, which relaxes the assumption that the intergenerational factor of the sibling effect is not a fixed proportion but varies across families, may better suit immigrant assimilation processes, given the documented role of family-specific factors in intergenerational persistence such as parental age at migration (Bleakley & Chin,2008), interethnic marriages (van Ours & Veenman,2010), intention to return to the country of origin (Dustmann, 2008), etc. In addition, ethnic capital might play a role in the persistence of inequalities across multiple generations (Borjas,1992;Gang & Zimmermann,2000). In particular, differences in average earnings between groups may persist over several generations because the performance of the offspring will also depend on source-country characteristics (Borjas,1993). Moreover, the results by Bingley & Cappellari (2019) may imply that some immigrants may take longer to assimilate, as intergenerational mobility may be lower than what is typically believed in the literature. In this paper, we exploit full-population Dutch administrative data across three generations to estimate sibling correlations in permanent income and to understand the relative weight of the intergenerational and the residual components in the sibling correlation in permanent earnings across different ethnic sub-groups. Our analysis relies on a model assuming a heterogeneous intergenerational income transmission process (Bingley & Cappellari,2019;Bingley et al.,2021), but we also provide results based on a model assuming homogeneity across families (Solon et al.,1991;Bj¨orklund et al.,2010). We then examine differences across immigration cohorts to study whether these change over time depending on the immigrant generation and explore heterogeneities based on income and ethnic segregation in the area of residence. We consider the native Dutch population and the three largest immigrant groups in the Netherlands, namely, those arriving from or having ancestors in: (i) European (EU-15) countries; (ii) the former Dutch colonies (Indonesia, Suriname, and the Antilles); (iii) Turkey and 3 Morocco.2While each of these groups accounts for roughly 5% of the current Dutch population, they differ substantially in terms of immigration history and assimilation process. We estimate comparable levels of sibling correlation across ethnic groups (0.298–0.315). Moreover, with respect to Dutch natives, immigrants of European and former-colonies origin seem to exhibit a higher degree of intergenerational transmission, while for immigrants from Turkey and Morocco this only accounts for half of the sibling correlation. We find a similar pattern when estimating IGC: while IGC of natives is 0.225, IGC of immigrants from Turkey and Morocco is only 0.163, and IGC of other immigrant groups ranges between 0.250 and 0.232.3 Next, we estimate sibling correlation by ethnic group and immigration generation separately, to understand whether the intergenerational transmission mechanism changes across generations, as immigrants become more assimilated to the host country. Indeed, the sibling correlation of first-generation immigrants is slightly lower than natives. Second-generation immigrants, on the other hand, align to natives. As for the role of parental income, we observe lower parental earnings transmission for all first-generation immigrant sub-groups. This gap, however, disappears when second-generation immigrants are compared to Dutch natives, with the sole exception of immigrants from Turkey and Morocco. Our results provide evidence of an assimilation process in the intergenerational transmission mechanism of immigrants but also highlight the existence of stark heterogeneities across immigrant groups. Thus, we further investigate potential channels to understand whether sibling correlations and the role played by the family component differ across community-specific characteristics. We exploit the richness of our data and compute measures of income and ethnic segregation based on the approach by Chetty et al. (2014) using information about the neighborhood of the parental household when the siblings were growing up. Our evidence implies that the relative importance of the parental component of the income transmission process and the immigrant assimilation dynamics are related to neighborhood characteristics like the degree of income and ethnic segregation. We observe that immigrants whose persistence of neighborhood status across generations is more pronounced follow a slower assimilation trajectory. For them, parental transmission no longer dominates in explaining sibling correlations in earnings, and this occurs especially at high levels of segregation. With this paper, we show the existence of heterogeneous patterns in the intergenerational transmission of permanent earnings across immigrants. To our knowledge, the only analysis using sibling correlation in the context of the immigrant sub-population is the one by Schnitzlein (2012), which relies on the homogeneity assumption and finds an average sibling correlation of 0.26 among second-generation immigrants in Denmark and of 0.17 among Danish natives. We extend our contribution to analysing sibling correlations following the recent approach by Bingley 2EU-15 countries are Belgium, Denmark, Germany, Finland, France, Greece, Ireland, Italy, Luxembourg, Austria, Portugal, Spain, the United Kingdom and Sweden. Turkey and Morocco are source countries of a substantial population of guest workers in the Netherlands. 3Sibling correlations are slightly larger than those estimated by Bingley & Cappellari (2019) in Denmark, but the share explained by parental income is in line with their estimates (between 50 and 90 percent). 4 & Cappellari (2019) and estimate the importance of the intergenerational factors in sibling correlations. We also add to the literature by looking into heterogeneities across immigrant generations and area-specific ethnic and income segregation levels. The remainder of this paper is structured as follows. In Section 2we briefly discuss the theoretical models in the literature. In Section 3we describe the recent immigration history of the Netherlands and the data used in the analysis. Section 4presents the main results, while Section 5reports heterogeneous results by immigrant generation and across neighborhoodspecific characteristics. Section 6concludes. 2 Intergenerational transmission processes Sibling correlations capture the influence of factors that siblings share in their socio-economic status (most often their earnings). These influences originate both from intergenerational factors and other components orthogonal to them. Models of sibling correlations assume log earnings in deviation from the mean (w) at time tfor individual iin family jas the sum of a permanent component (y) and an orthogonal, white noise, transitory shock (v): wijt =yij +vijt, vijt ∼N(0, σ2 v) (1) The permanent component is then divided into two parts: fjrepresents the component shared by the siblings and aij captures the individual-specific deviation from fj(i.e. a component unique to the individual i): yij =aij +fj, aij ∼N(0, σ2 a), fj∼N(0, σ2 f) (2) Given that the family component fjand the individual-specific component aij are orthogonal, in the long run the income variance is the sum of their variances (Solon et al.,1991). Therefore, the sibling correlation can be explained as: rS=σ2 f σ2 f+σ2 a .(3) Here, rScan be thought of as a comprehensive measure of family background and community effects (Bj¨orklund et al.,2002;Mazumder,2011;J¨antti & Jenkins,2015). It contains intergenerational components, such as parental income, cultural inheritance, and parental-driven community effects, as well as individual-specific components, e.g., genetic traits not shared by the siblings, time-specific community effects, and differences in parental involvement in the offspring’s education. Not all family and community effects are captured by the sibling correlations. Indeed, all factors not shared by the sibling, such as changes across time in school and neighborhood quality, differences in parental involvement in the offspring’s education, and idiosyncratic genetic traits are not captured by sibling correlations. Thus, sibling correlations 5 may underestimate the importance of family background and community effect (Bj¨orklund & J¨antti,2020). Estimating the degree to which rSis driven by intergenerational components does not only help to understand the channels behind between-family differences but also the role of other factors such as neighborhood and network effects (Benabou,1994;Durlauf,1994;Chetty et al.,2014; Chetty & Hendren,2018a). The intergenerational component is usually estimated either through ANOVA (Solon et al.,1991;Bj¨orklund et al.,2002), generalized method of moments (Bj¨orklund et al.,2009) or mixed models via restricted maximum likelihood (Levine & Mazumder,2007; Mazumder,2008;Bj¨orklund et al.,2010). These models have different underlying assumptions in terms of stationarity and the form of the distribution of the unobserved components. However, they all suggest that parent-to-child transmission is not the main component of the sibling correlation, which is to say that rSis mostly driven by factors that siblings share other than parental earnings.4Further, they all share the underlying idea that the intergenerational transmission process is homogeneous across families. Bingley & Cappellari (2019) relax such homogeneity assumption and argue that the role of the intergenerational component is much higher than what previous studies usually find. Using Danish administrative data, they estimate a sibling correlation of 0.22 and that this is largely explained by the intergenerational transmission process (by 72%). This is in contrast with previous studies, and in particular with the estimates for Denmark by Bj¨orklund & J¨antti (2009), where the intergenerational component accounts for only 6% of the sibling correlation. As in Equation 1, the model of Bingley & Cappellari (2019) assumes that wijt consist of a permanent and a transitory component, which are orthogonal, such that E(yij +vijt) = 0. However, instead of assuming that the intergenerational component fjis homogeneous across families, they allow it to vary across families, so that fj=µI j+µR j, where µI jis the family-specific intergenerational component and µR jis the residual component capturing those attributes shared by the siblings which do not depend on fathers’ earnings. Therefore, they write Equation 2as: yij =aij +µI j+ +µR j,∀i:m(i) = S1, S2, aij ∼N(0, σ2 am), µR j∼N(0, σ2 µR), µI j∼N(0, σ2 µI )(4) for the first-born (S1) and second-born (S2) sons and as: yij =aij +µI j, aij ∼N(0, σ2 am),∀i:m(i) = F(5) for the fathers, which do not have any residual sibling component µR j. In both Equations 4and 5, the variance component allows for family-member mspecific distributions of the individual-specific components aij. Modeling fathers’ earnings jointly with the firstand second-born sons in a Generalized 4See Solon et al. (1991); Bj¨orklund et al. (2004) and J¨antti & Jenkins (2015) for a survey on the evidence on sibling correlations. 6 Figure 2: Maps (a) Share of immigrants (b) Average income (c) Income segregation (d) Ethnic segregation Note: Geographical distribution of (a) the share of immigrants, (b) average income, (c) income segregation, and (d) ethnic segregation. The first two variables are measured at the neighborhood level, while segregation indexes are computed at the municipality level. This province encompasses the Dutch seat of government, The Hague, alongside Rotterdam, its largest urban center. Additionally, the Limburg region, situated as the southernmost of the twelve provinces, is also marked by a notable immigrant presence, with Maastricht as its provincial capital. Amsterdam, the national capital, assumes prominence by hosting a substantial international community, a phenomenon attributed to its historical eminence and cosmopolitan ambiance. The geographical triangle delineated by the provinces of South Holland, North Holland, and Utrecht, which encompass major urban centers such as Amsterdam, Rotterdam, The Hague, and Utrecht, are associated with greater economic activity and opportunities, contributing to elevated average income levels shown in panel Figure 2b. Finally, Figures 2c and 2d reveal a pervasive dispersion of both income and ethnic segregation across the Netherlands. However, there are some spatial concentrations in close proximity to major urban agglomerates 13 Table 2: Main results (a) GMM Dutch EU-15 Ex-Colonies Turkey-Morocco Coeff. S.E. Coeff. S.E. Coeff. S.E. Coeff. S.E. rS0.312 (0.003) 0.315 (0.010) 0.309 (0.010) 0.298 (0.012) rI/rS0.830 (0.022) 0.923 (0.047) 0.858 (0.040) 0.538 (0.033) IGC 0.225 (0.006) 0.250 (0.010) 0.232 (0.008) 0.163 (0.008) (b) REML Dutch EU-15 Ex-Colonies Turkey-Morocco Coeff. S.E. Coeff. S.E. Coeff. S.E. Coeff. S.E. rS0.258 (0.002) 0.249 (0.006) 0.250 (0.006) 0.242 (0.007) rI/rS0.084 (0.002) 0.105 (0.008) 0.130 (0.008) 0.093 (0.010) IGC 0.147 (0.002) 0.162 (0.006) 0.180 (0.005) 0.150 (0.008) Note: Panel (a) reports the results of GMM estimates; panel (b) those of REML estimates. rSis the sibling correlation in income; rIis the part of rSthat is due to intergenerational components; rI/rSis the share of the sibling correlation that is due to intergenerational components. IGC is the intergenerational correlation in income (father-to-child correlation). and the capital. 4 Results across ethnic groups We study the patterns in sibling correlation in permanent earnings (rS) across immigrant groups and further decompose it to estimate the extent to which this is due to intergenerational components (i.e. due to parental permanent income, rI/rS). We also provide estimates of the father-to-child correlation (or intergenerational correlation, IGC). Our preferred model is the one assuming that intergenerational transmission in income is heterogeneous across families, following Bingley & Cappellari (2019). We show results for Dutch natives, which we take as a benchmark, immigrants originating from EU-15 countries, former colonies, and Turkey and Morocco separately.11 Table 2reports our main results. Our GMM estimates of the sibling correlation are similar across ethnic groups and range between 0.298 and 0.315. Importantly, the patterns are in line with the estimates for the group of Dutch natives, for whom the sibling correlation is 0.312. In comparison to the Danish context, our estimates of the sibling correlation are slightly larger than the 0.22 estimate by Bingley & Cappellari (2019). A similar result emerges also when estimating the sibling correlation with a model assuming that intergenerational transmission in income is homogeneous across families (REML estimates). Indeed, while point estimates are slightly smaller than with GMM, the 11 Due to the difference in the timing of arrival shown in Figure 1, we also consider immigrants from Indonesia and Suriname separately. The results are identical. 14 sibling correlation across ethnic groups ranges from 0.242 for the Turkey-Morocco immigrant group to 0.258 for the Dutch natives. Thus, differently from Schnitzlein (2012), who estimates sibling correlations equal to 0.165 for Danish natives and to 0.263 for second-generation immigrants, and regardless of the model used, we do not observe substantially larger values of rSfor immigrants than for natives. Next, we investigate whether the fact that sibling correlation in income is very similar across ethnic groups also implies that the relevance of parental permanent income is comparable. Thus, we estimate the intergenerational share, i.e. the ratio between the intergenerational component rIand the sibling correlation rS. As reported in Table 2a, we observe quite high values for both European immigrants and those from former colonies (0.923 and 0.858, respectively). These are higher than that referred to Dutch natives, for whom 83% of the sibling correlation is attributable to the intergenerational component rI. Conversely, for individuals originating from Turkey and Morocco, rIonly contributes to around half of the sibling correlation. These results confirm those by Bingley & Cappellari (2019), as heterogeneously distributed intergenerational transmission accounts for a large fraction (between 50 to 90 percent) of the overall sibling correlation of permanent earnings. Moreover, our estimates indicate that the subgroups of immigrants who are more similar to Dutch natives due to historical and institutional reasons (i.e. Europeans and those from former colonies) also tend to have comparable patterns in parental influence. Other immigrants, including guest worker immigrants –typically sourced from Turkey and Morocco–, while displaying analogous levels of sibling correlation, owe half of this to factors independent of parental income. As they have lower average levels of income (see Table 1), it seems reasonable to expect that their higher degree of mobility translates into less transmission of poverty across generations. When we allow for homogeneity in intergenerational transmission between families, the coefficient referred to individuals of former colonies’ origin reaches 0.130, while it is around 0.100 for the other immigrant groups and 0.084 for natives (Table 2b). This points to the existence of fairly similar intergenerational patterns across all groups. However, this might be due to the fact that estimates in the REML model are relatively compressed, thus also cross-group differences are less noticeable. Finally, we consider the intergenerational correlation in income (IGC), or father-to-child correlation, across ethnic groups. In comparison to a point estimate of 0.225 for Dutch natives, we find that immigrants tend to have slightly higher values (see Table 2a). As before, immigrants from Turkey and Morocco are an exception, as this group seems to experience a relatively less important role of parental influence (0.163). Such difference might be explained by Turkish and Moroccan immigrants starting from a lower socio-economic position compared to Dutch natives and to Europeans and those coming from former colonies (van Ours & Veenman,2003). Hence, it is reasonable for them to experience a higher degree of mobility. Such a pattern, although less evident, is confirmed under the REML model (Table 2b). 15 Figure 3: Results by immigrant generation (a) Sibling correlation (b) Intergenerational share Note: Each plot reports point estimates and corresponding confidence intervals at 95% level. Panel (a) shows results on the sibling correlation in income (rS); panel (b) those on the share of the sibling correlation that is due to intergenerational components (rI/rS). Point estimates and standard errors are also reported in Table A.2. 5 Results across generations and neighborhood characteristics In the previous section, we show that Dutch natives and immigrants have comparable levels of sibling correlation. Yet, for almost all ethnic groups the intergenerational component and the father-to-child correlation are higher than those of natives. Estimates for immigrants of Turkish and Moroccan descent are considerably lower, suggesting a less relevant role of the intergenerational component (i.e., parental earnings) and a more influential role of the residual component in explaining the sibling correlation. These differences in intergenerational transmission dynamics across ethnic groups might hide heterogeneous patterns in the assimilation mechanism over immigrant generations. Here, we explore the existence of separate paths across sub-groups. When addressing the evolution of intergenerational transmission in income for immigrants, we might expect that ethnic disparities fade as families settle and acquire host country-specific human and social capital. Thus, second-generation immigrants would behave more similarly to natives, and so would their intergenerational transmission mechanism. Figure 3a reports our estimates of the sibling correlation in permanent earnings by immigrant generation. Point estimates are fairly similar across generations and ethnic groups, although those referred to as first-generation immigrants are slightly lower than those of natives.12 When we compute intergenerational shares by immigrant generation (Figure 3b), however, estimated parameters suggest a large heterogeneity in the contribution of the intergenerational component to the sibling correlation across sub-groups. In particular, the role of parental permanent income seems to be much higher for natives, while the residual component has a substantial role for 12 Coefficients and standard errors are also reported in Table A.2. We exclude the group of first-generation EU-15 immigrants due to its small sample size (i.e., less than 600 observations). 16 first-generation immigrants. When considering second-generation immigrants, estimated values align with the ones referred to by the Dutch natives. The only exception is given by immigrants of Turkish and Moroccan origin, who maintain the same degree of importance of the intergenerational component as the previous generation. These results seem to confirm the hypotheses mentioned above. If the intergenerational transmission mechanism tends to assimilate across generations, this might be because the characteristics of the parental and residual components become similar to those of natives and so does their relative weight in explaining the sibling correlation. Our estimates suggest that this process occurs within the span of just two generations for immigrants from EU countries and former Dutch colonies, possibly due to the shared history and institutions with the native population. On the other hand, the ethnic group that does not share such strong ties with the Dutch native population seems to experience a slower assimilation process. We further investigate the potential reasons behind these heterogeneities. Over the last decades, a large literature has investigated the drivers of intergenerational mobility differences and the role of mediating factors such as neighborhood and school quality (B¨ugelmayer & Schnitzlein,2018;Hedman et al.,2019;Hedman & Ham,2021). Above all, neighborhood effects have been proven to be crucial for long-term outcomes such as earnings, via their temporal dimension (Chetty & Hendren,2018a). Thus, the differences in intergenerational mobility between natives and immigrants of different ethnic backgrounds uncovered in Figure 3might mask some heterogeneities deriving from these long-term exposures. Indeed, Kleinepier et al. (2018) find that non-Dutch children, especially Turkish and Moroccan, are more likely to live in poor neighborhoods than their native counterparts, not only at a specific point in time during their childhood but also throughout their entire youth.13 We explore the role of neighborhood characteristics and consider the degree of segregation in income HIand ethnic segregation HE(Chetty et al.,2014). As described in Section 3.2, we define as highly segregated those neighborhoods that belong to the top quartile of the distribution. Figure 4reports the share of individuals living in highly segregated neighborhoods by ethnic origin and immigrant generation. Three considerations arise. First, the two measures, which are strongly correlated, give very similar distributions. Second, the proportion of individuals living in highly segregated neighborhoods is extremely high among first-generation immigrants (around 70%), and significantly lower among Dutch natives (around 23%). Third, the proportion of individuals living in highly segregated neighborhoods reduces substantially for second-generation immigrants, except for those of Turkish and Moroccan origin. Together with the evidence in Figure 3this suggests a link between assimilation patterns, the degree of income and ethnic segregation and the relative weight of the parental component of the income transmission process. In particular, neighborhood characteristics might be relatively less persistent for the groups that quickly assimilate the intergenerational dynamics of the native 13 With Swedish data, Manley et al. (2020) also find that childhood neighborhood has long-lasting effects on residential careers, but these are attenuated by family effects. 17 Figure 4: Community characteristics by immigrant generation (a) Highly segregated in income (b) Highly segregated in ethnicity Note: Each plot reports the share of individuals living in a highly segregated neighborhood by ethnic origin and immigrant generation. Panel (a) refers to income segregation (HI); panel (b) to ethnic segregation (HE). population. Conversely, immigrants on a slower assimilation path, for whom the transmission of neighborhood status is more persistent across generations, seem to experience a lower influence of the parental transmission into their sibling correlation in income. This suggests that neighborhood dynamics might contribute to explaining the intergenerational transmission mechanism for the latter immigrant group. Thus, we turn to directly assessing the importance of living in ethnically or economically segregated areas by considering ethnicand income-based segregation in the municipality as measures of spatial heterogeneity. Figures 5and 6report the results by level of income and ethnic segregation in the neighborhood, respectively. “Low” and “High” refer to neighborhoods in the bottom three quartiles and the top quartile of the distribution, respectively, as described in Section 3.2. Estimated sibling correlations in earnings are in line with our main results, regardless of the level of segregation in the neighborhood of residence, and so is the estimated intergenerational share. As expected, for immigrants from Turkey and Morocco the estimated intergenerational component is overall relatively less important, but it is especially low the higher the level of segregation. Indeed, those residing in highly segregated areas tend to exhibit below-average values of the intergenerational share, suggesting that this dimension is not entirely captured by the intergenerational component explaining the sibling correlation in permanent income. This result is especially relevant because it implies that the importance of the neighborhood is not homogeneous across neighborhood characteristics. Moreover, it shows that neighborhood effects influence transmission especially for the ethnic group experiencing higher, and more persistent, segregation levels. In the Appendix, as robustness, we provide additional results based on the poverty seg18 Figure 5: Results by income segregation HI (a) Sibling correlation (b) Intergenerational share Note: Each plot reports point estimates and corresponding confidence intervals at 95% level. Panel (a) shows results on the sibling correlation in income (rS); panel (b) those on the share of the sibling correlation that is due to intergenerational components (rI/rS). Point estimates and standard errors are also reported in Table A.2. Figure 6: Results by ethnic segregation HE (a) Sibling correlation (b) Intergenerational share Note: Each plot reports point estimates and corresponding confidence intervals at 95% level. Panel (a) shows results on the sibling correlation in income (rS); panel (b) those on the share of the sibling correlation that is due to intergenerational components (rI/rS). Point estimates and standard errors are also reported in Table A.2. regation index HP(Figure A.1) and on neighborhood-specific measures of income and ethnic heterogeneity. We consider the share of households having income below the 25th percentile (Figure A.2) and the share of non-Dutch residents (Figure A.3). Estimates yield similar conclusions as those presented above. 19 6 Conclusions For decades, the variation in earnings shared by siblings was believed to be mostly due to factors uncorrelated with parental earnings (Solon et al.,1991;Bj¨orklund et al.,2010). The recent contribution by Bingley & Cappellari (2019) shows that when one assumes heterogeneity in intergenerational transmission between families, intergenerational persistence accounts for more than two-thirds of the sibling correlation in permanent earnings. It is not trivial, however, whether this new evidence applies also to the immigrant population and to what extent it may explain differences in immigrant assimilation paths across groups and generations. We follow the approach of Bingley & Cappellari (2019) to decompose the sibling correlation of permanent earnings into intergenerational and residual sibling components and to compute the corresponding intergenerational share for a sample of Dutch and immigrant individuals over three generations. We find that a substantial fraction of the overall sibling correlation, indeed, derives from parental influences. When considering sub-populations on the basis of ethnic origin, the sibling correlation due to intergenerational transmission is similar to that of natives for immigrants from Europe and former Dutch colonies, while it is sensibly lower for those originating from Turkey and Morocco, who moved to the Netherlands mainly as guest-workers and share different characteristics with respect to Dutch natives and other immigrants. When we analyze individuals by immigrant generation, two important results emerge. First, sibling correlation in earnings is slightly lower for first-generation immigrants compared to second-generation ones, who tend to align with Dutch natives. Second, the weight of parental income in the sibling correlation in earnings is lower for immigrants compared to natives. It returns to values comparable to the native population for second-generation immigrants from Europe and former colonies, while for immigrants of different origins (Turkey and Morocco) factors unrelated to parental income continue to have a meaningful role. These heterogeneities suggest that different assimilation mechanisms might be at work. Thus, we follow the literature in exploring the potential role of mediating factors that could explain such discrepancies across groups (B¨ugelmayer & Schnitzlein,2018;Chetty & Hendren,2018b; Hedman & Ham,2021). We consider two spatial measures of neighborhood ethnic and income segregation. Our analysis reveals that the patterns observed for the estimated intergenerational share also arise when considering the share of individuals living in highly segregated neighborhoods. We also find that that the role of the intergenerational component in explaining the sibling correlation in earnings is especially low for the group of immigrants for whom the transmission of neighborhood status persists across generations. We conclude that, while the sibling correlation in income is fairly similar across native and immigrant groups, this correlation does not appear to be explained by the same factors. As immigrants assimilate into the native population, their intergenerational transmission mechanisms also become similar to the ones experienced by natives. However, such an assimilation mecha20 nism is not homogeneous across groups. Our results imply that neighborhood effects are more relevant, and persistent, for those immigrants who are on a slower assimilation path experience. 21 References Algan, Y., Dustmann, C., Glitz, A., & Manning, A. (2010). The economic situation of first and second-generation immigrants in France, Germany and the United Kingdom. Economic Journal,542 , F4–F30. Becker, G. S., & Tomes, N. (1976). Child endowments and the quantity and quality of children. Journal of political Economy,84 , S143–S162. Becker, G. S., & Tomes, N. (1979). An equilibrium theory of the distribution of income and intergenerational mobility. Journal of Political Economy,87 , 1153–1189. Benabou, R. (1994). Human capital, inequality, and growth: A local perspective. European Economic Review,38 , 817–826. Bingley, P., & Cappellari, L. (2019). Correlation of brothers’ earnings and intergenerational transmission. Review of Economics and Statistics,101, 370–383. Bingley, P., Cappellari, L., & Tatsiramos, K. (2021). Family, community and long-term socioeconomic inequality: Evidence from siblings and youth Peers. Economic Journal,131 , 1515– 1554. Bj¨orklund, A., Eriksson, T., J¨antti, M., Raaum, O., & ¨ Osterbacka, E. (2002). Brother correlations in earnings in Denmark, Finland, Norway and Sweden compared to the United States. Journal of Population Economics,15 , 757–772. Bj¨orklund, A., Eriksson, T. V., J¨antti, M., Raaum, O., & ¨ Osterbacka, E. (2004). Family structure and labor market success: The influence of siblings and birth order on the earnings of adults in Norway, Finland, and Sweden. In Generational Income Mobility in North America and Europe. Cambridge University Press. Bj¨orklund, A., & J¨antti, M. (2009). Intergenerational income mobility and the role of family background. Oxford Handbook of Economic Inequality,491, 521. Bj¨orklund, A., & J¨antti, M. (2020). Intergenerational mobility, intergenerational effects, sibling correlations, and equality of opportunity: A comparison of four approaches. Research in Social Stratification and Mobility,70, 100455. Bj¨orklund, A., J¨antti, M., & Lindquist, M. J. (2009). Family background and income during the rise of the welfare state: Brother correlations in income for Swedish men born 1932–1968. Journal of Public Economics,93 , 671–680. Bj¨orklund, A., Lindahl, L., & Lindquist, M. J. (2010). What more than parental income, education and occupation? An exploration of what Swedish siblings get from their parents. BE Journal of Economic Analysis & Policy,10. 22 Getting in touch with the EU In person All over the European Union there are hundreds of Europe Direct centres. You can find the address of the centre nearest you online (european-union.europa.eu/contact-eu/meet-us_en). On the phone or in writing Europe Direct is a service that answers your questions about the European Union. You can contact this service: — by freephone: 00 800 6 7 8 9 10 11 (certain operators may charge for these calls), — at the following standard number: +32 22999696, — via the following form: european-union.europa.eu/contact-eu/write-us_en. Finding information about the EU Online Information about the European Union in all the official languages of the EU is available on the Europa website (european-union.europa.eu). EU publications You can view or order EU publications at op.europa.eu/en/publications. Multiple copies of free publications can be obtained by contacting Europe Direct or your local documentation centre (european-union.europa.eu/contact-eu/meet-us_en). EU law and related documents For access to legal information from the EU, including all EU law since 1951 in all the official language versions, go to EUR-Lex (eur-lex.europa.eu). EU open data The portal data.europa.eu provides access to open datasets from the EU institutions, bodies and agencies. These can be downloaded and reused for free, for both commercial and non-commercial purposes. The portal also provides access to a wealth of datasets from European countries. The Joint Research Centre (JRC) provides independent, evidence-based knowledge and science, supporting EU policies to positively impact society EU Science Hub Joint-research-centre.ec.europa.eu