Introduction

Over the past two decades, the theoretical concept of “immigrant selectivity” has received widespread attention from social researchers. This concept refers to how immigrants differ from individuals in their society of origin who did not migrate. Its theoretical significance, however, lies in explaining such outcomes of immigrant incorporation in the receiving society as health and labor market incorporation, as well as their children’s educational and occupational achievements. In general, research has demonstrated that immigrants tend to be positively selected on indicators such as education, income, skill, and health compared to their non-immigrant counterparts at home (e.g., Engzell & Ichou, 2020; Feliciano, 2005a; Ichou, 2014; Ton & Haris, 2021; Welker & Will, 2023; van de Werfhorst & Heath, 2019). This pattern, in turn, may explain “paradoxical” findings suggesting advantages associated with being an immigrant or the offspring of immigrants compared to individuals from the receiving society (for a recent review of theory and findings, see Feliciano, 2020).

Despite a large body of research on immigrant selectivity, little attention has been given to its temporal dimension, meaning how immigrant social characteristics change over time. Thus, while the phenomenon of immigrant selectivity has been theoretically identified and empirically demonstrated, it has been viewed as a relatively stable social occurrence. Few studies have examined changes in immigrant selectivity over relatively short periods of time (Feliciano, 2018; Jones, 1998; Lindstrom & López Ramírez, 2010; Massey et al., 1990), leaving us with limited knowledge of how the characteristics of immigrants from the same country of origin to the same destination country change during an ongoing wave of immigration (Jones, 1998; Massey et al., 1990; Popkin, 1999; Rhoades, 1978). Our study addresses this gap by analyzing data on Jewish immigrants to Israel from the former Soviet Union (FSU) between 1989 and 1994. We focused on whether the educational attainment of immigrant parents (in their country of origin) and their children (in the host country) remained constant or decreased with the year of arrival and whether changes in parental education or other sociodemographic characteristics could explain any changes in the children’s educational achievements. By examining changes over a short period, we could assume that there were no substantial changes in the average level of education of the general population of the country of origin during this time frame.

Literature Review

Immigrant Selectivity and Children of Immigrants’ Educational Outcomes

Immigrant selectivity has received considerable attention from social scientists in the past two decades. This body of research suggests that positive immigrant selectivity can help account for “paradoxical” patterns of success among immigrants and their children in education, labor markets, and health. In her recent review, Feliciano (2020) concludes that immigrant selectivity is an important explanation for two salient patterns found in educational research on migration: (1) considerable variability in the association of social background and educational outcomes among children of immigrants; and (2) better educational outcomes of children with migration background compared with non-migrant children with similar socioeconomic background.

These patterns were found in several migrant-receiving countries. For example, Feliciano’s studies (2005a, 2005b, 2006, 2008, 2018) provide evidence that most immigrant groups in the USA are positively selected on education, and variations in academic selectivity may account for group differences in educational expectations and actual attainment levels. She also found that women tend to be more selective in terms of education than men among Mexican immigrants. Tong and Harris’ study (2021) on immigrants to the USA from Asian countries found that parental educational attainment in the country of origin predicted children’s chances of college enrollment in the destination country. Ichou’s study (2014) on immigrants in France found variability in educational selectivity both between and within groups from different countries of origin, and parental relative educational attainment in the country of origin had a substantial positive effect on children’s educational attainment in the destination country. Engzell and Ichou’s study (2020) using Swedish data found a positive effect of parental relative educational attainment on immigrant children’s aspirations, attitudes, and placement in an academic track in high school. Van de Werfhorst and Heath’s study (2019) on second-generation immigrant students from a large number of origin countries in Canada, the USA, and eight European countries found that the disadvantage of second-generation students compared to their majority counterparts was smaller for immigrant groups characterized by high favorable educational selectivity. Their study also suggests that the effect of immigrant education selectivity is moderated by receiving countries’ integration policies and educational policies related to high school tracking arrangements.

Overall, this international body of research provides compelling evidence that immigrant selectivity plays a crucial role in shaping the educational outcomes of immigrant children in destination countries. They also highlight the importance of social and education policies in understanding and addressing the effects of immigrant selectivity on educational outcomes. Further research is needed to explore the mechanisms through which immigrant selectivity operates and to identify effective policies to promote positive outcomes for immigrant children.

Stages in the Immigration Stream and Decreased Positive Selectivity over Time

As our above review shows, recent literature on immigrant selectivity largely overlooks changes over time in immigrant selectivity. This is surprising, as we know migration is a dynamic process: it constantly evolves to reflect macro changes in both countries of origin (e.g., political crises or stability, wars, economic turns, climate conditions such as droughts) and countries of destination (e.g., immigration policy, economic turns, public opinion). Beyond these external macro effects on immigration streams, scholars have identified several internal dynamics and processes in creating immigrant communities. These include, for example, forming networks that carry information on opportunities and obstacles in the destination country and reducing immigration costs over time (e.g., De Haas, 2010, 2019; Haug, 2008; Massey, 2020).

To reflect this evolving process, researchers have developed a theoretical distinction between the initial, developmental, and mature stages of immigration flow. In the initial stage, “pioneers” arrive in the receiving country, establish information and support networks, and encourage the arrival of others from the same country of origin. In the developmental stage, immigration changes from an individual to a social process based on social networks, and this helps lower barriers to immigration. Consequently, the positive selectivity of migration declines from the initial to the developmental stage. As the stream of immigration continues, in the third stage, a community of immigrants from the same origin is established; these immigrants “are representative of the local populations from which they are drawn and may even become negatively selected” (Lindstrom & López Ramírez, 2010, p. 56).

In an early study drawing on data from Monterey, Mexico, Browning and Feindt (1969) examined immigration patterns to this city over time. They found immigrants were positively selected on several attributes, including the level of education, compared to the populations from which they originated. Still, they had become less selective over time. Quantitative studies by Massey et al. (1990) and Jones (1998) showed similar patterns for immigration from Mexico to the USA. Qualitative studies by Rhoades (1978) on immigration from southern Spain to Germany and by Popkin (1999) on Guatemalan immigrants to the USA further developed the theoretical argument about decreased immigrant selectivity over time. More recently, Lindstrom and López Ramírez (2010) used data from the USA to explore differences between male “pioneers” and “followers” in immigration from several Latin American countries (Costa Rica, Dominican Republic, Guatemala, Mexico, and Nicaragua). They relied on the transition from the initial to the developmental stage to distinguish between these two groups (but did not distinguish between immigrants in the second and third stages). Pioneers who arrived in the initial stage were younger and less likely to be married than followers who arrived in the developmental stage. They also were less likely to own a business or agricultural land before migrating. Of interest to our study, the level of education did not differ for pioneers and followers in most origin groups. Still, both pioneers and followers had a higher average level of education than those who remained in their country of origin.

Feliciano (2008), who studies immigration from Mexico to the USA, found that educational selectivity increased for both men and women, especially between 1960 and 1990. Feliciano suggests that the increased selectivity among Mexican migrants, in contrast to the theoretical expectation based on network expansion and reduced costs of migration, may result from changes in the nature of migration from Mexico. She further explains that in the past, most Mexican migrants came from rural areas, while in more recent decades, growing numbers originated from urban areas. This demographic trend can explain increased levels of education among Mexican migrants. Since her data did not allow for a distinction between rural and urban areas of origin, she was unable to examine temporal trends in selectivity among migrants from rural areas and among migrants from urban areas.

The relatively scarce research on the temporal dimension of immigrant selectivity and its inconsistent findings call for further exploration of this topic. In the present paper, we aim to contribute to bridging this gap in the literature by analyzing data on migration from the FSU to Israel in the early 1990s. By limiting the analysis to five years, we can examine theoretical arguments about declining immigrant selectivity without the confounding effects of demographic changes that may occur in the country of origin over longer periods of time.

Educational Attainment of Children of FSU Immigrants

Following the collapse of the FSU, many of its citizens with Jewish ancestry have immigrated to Israel under the Israeli Law of Return. From 1989 and onwards, hundreds of thousands of people left the FSU, heading for Israel, with about one million arriving in the first decade of this ongoing immigration wave. FSU immigrants currently comprise about 15% of the country’s total population and about 20% of its Jewish majority. Given this massive influx of migrants, the Israeli Ministry of Education and other public institutions had made considerable efforts to develop mechanisms for absorption that prioritize language acquisition, social and educational integration in schools, and informal education. For a comprehensive overview of the cultural and socio-demographic characteristics of this migration wave, as well as social and educational policies to meet the needs of immigrants during the 1990s, and small-scale studies on the incorporation of children and youth into Israeli society and its education system, see Horowitz (1999).

While there is relatively limited quantitative research on the educational achievements of FSU immigrants in the Israeli education system, the findings reveal complex and sometimes contradictory patterns. For example, Levin and Shohamy (2008) compared the Hebrew and mathematics achievements of FSU immigrant students with those of native-born students. Focusing on the fifth, ninth, and eleventh grades, they found that FSU immigrant students performed more poorly in both subjects than non-immigrant students (except for mathematics in the ninth grade). Feniger (2017) used data from the international PISA 2006 study to compare school-related attitudes, aspirations, and science achievement among 15-year-old immigrants from the FSU and native-born Israeli teenagers. The study revealed that FSU immigrants had, on average, higher scores in a standardized science test and higher aspirations for science-related careers in the future than their native-born comparators.

Chachashvili-Bolotin’s research in 2007 and 2011 delved into high school completion and high school matriculation diplomas for FSU immigrants. Their analysis of a large representative sample of Israelis who graduated from high school in the mid-1990s found that FSU immigrants had a polarized pattern of behavior. While they were more likely to drop out than native-born Israelis, those who remained (the majority of the sample) were more likely to attain an advanced matriculation diploma. Additionally, FSU immigrants specialized in either less-selective vocational tracks or highly selective academic scientific subjects in high school. This specialization in the vocational track exemplifies the instrumentalist approach FSU immigrants take to education. These tracks may be less demanding than humanities, but they promise specific occupational skills useful to new labor market entrants immediately after high school.

Regarding enrollment in higher education, Feniger et al. (2015, 2023), utilizing large representative samples, found that FSU immigrants who qualified for the matriculation diploma were less likely to pursue higher education compared to their non-immigrant peers. However, those who did opt for academic pursuits were more likely than native-born Jews to enroll in lucrative fields of study, leading to increased economic mobility. These patterns could potentially be attributed to changes over time within the wave of FSU immigration to Israel. By disaggregation of the migration wave by year, valuable insights into changes in educational achievement and immigrant selectivity could be uncovered.

Study Aims and Research Questions

Interestingly, while research on immigration selectivity has flourished in the last two decades, arguments on the possibility of decreased positive selectivity over time as immigrants from a particular destination country that continue to arrive have been largely overlooked, despite what earlier studies suggested. We bridge this gap by focusing on arriving immigrants and their children’s educational attainment in the destination country. Our main theoretical argument was that while immigrant selectivity is crucial for understanding immigrant outcomes in the receiving country, it is equally important to consider temporal changes in this selectivity. We test this theoretical argument by examining immigration from the FSU to Israel between 1989 and 1994. The case of FSU immigrants in Israel offers a unique opportunity to study the question of decreased selectivity, as it concerns a large wave of immigration from one country of origin to one country of destination in a limited period. The limited period allows us to neutralize, to a certain extent, the effects of demographic changes in the country of origin that might occur over a longer period of time. It thus helps overcome two main limitations hindering previous studies: (1) small samples that limit the ability to distinguish between different years or stages in the stream of immigration from a single country of origin; and (2) exogenous effects on both countries of origin and destination if the study considers an extended period.

Using census and administrative data, we analyzed a large sample of FSU immigrants and their children, focusing on the level of educational attainment for both parents (in the FSU) and children (in Israel). Our primary research questions were as follows:

  1. 1.

    Does the educational attainment of immigrant parents remain constant or decrease with the year of arrival?

  2. 2.

    Do the educational achievements and attainments of children of immigrants remain constant or decrease with the year of arrival?

  3. 3.

    If the answer to question 2 is positive, can it be explained by changes in parental education and other sociodemographic characteristics?

Data and Variables

Sample

The Israeli Central Bureau of Statistics (CBS) prepared the dataset for this study that covers cohorts born between 1978 and 1985. It combines data from the 1995 Israel population census with newer data from the Ministry of Education and tertiary education institutions. The data on parental characteristics refer to 1995, and the information on children’s educational outcomes (in adolescence and adulthood) is updated to 2017. The dataset is based on information drawn from a representative sample of 20% of the Israeli households participating in the 1995 national census (for more information on sampling in the 1995 national census, see Hleihel, 2006). We focused on FSU immigrant children arriving in Israel between 1989 and 1994 who studied in Hebrew-speaking state and state-religious schools. We restricted the analysis to those who came to Israel between the ages of six and twelve (i.e., 1.5-generation; Rumbaut, 2004) during the elementary stage of education. Since our data cover those born until 1985, only a few were under six when they arrived in Israel. A few immigrant students attending ultra-Orthodox independent schools were omitted from the study because their curricula differ markedly from mainstream state and state-religious schools. In line with the strict rules of the Israeli CBS aimed at securing the confidentiality of the data, analysis was carried out in a secured room within the CBS offices.

Dependent Variables

Eligibility for a Matriculation Diploma at the End of High School

In Israel, obtaining a matriculation diploma is a crucial achievement after completing secondary education. This diploma is a prerequisite for higher education and opens up various labor market opportunities. The Ministry of Education administers the matriculation examinations during the 10th, 11th, and 12th grades based on study units. The number of study units ranges in a specific subject from 2 to 5. These units are equivalent to three weekly hours per year or one weekly hour per three years, and in most subjects, exams can be taken at either basic or advanced levels (depending on the number of study units taken). To receive a matriculation certificate, students must earn at least 20 units and take exams in all compulsory subjects, including Civics, Hebrew, Literature, and Bible (for the Jewish sector of education, which is the focus of this study), as well as at least one advanced subject. For more information on Israeli secondary education and the matriculation exams, see Ayalon (2006) and Bar-Haim and Feniger (2021).

Score on the Hebrew Matriculation Exam

Hebrew is a compulsory subject in the Israeli matriculation diploma, and all students are expected to take it. Immigrant students take a particular adjusted exam that differs from the regular exam. In addition to being part of the requirements of the matriculation diploma, the Hebrew exam measures a crucial skill for new immigrants when they enter the Israeli higher education system and labor market—proficiency in the host country’s primary language (see, e.g., Feniger et al., 2023). The original range of scores was 0–100, but in the dataset we used, they were recoded (for reasons of confidentiality) into six categories: 91–100, 81–90, 71–80, 61–70, 55–60, 54, or less (a failure). In order to convert this variable into a continuous variable that enables linear regression analysis, we assigned a value representing the middle of the category for each category. About 34.5% of the students in our sample did not take the test (similar to the equivalent percentage in the non-immigrant Jewish population), either because they dropped out of high school before taking the test or because their school decided not to have them registered for the test because of poor achievement. In the analysis, they were assigned a score of 0. Our preliminary analysis showed that the percentage of those who did not take the test was relatively stable for the different years of migration arrival. Yet, in more recent years (1992, 1993, 1994), it slightly increased. We conducted a robustness analysis to account for this trend, as shown in the Appendix.

Attainment of an Academic Degree

This was a dummy variable (1 = eligibility for an academic degree). In Israel, as in many other countries, attaining an academic degree is associated with important advantages in the labor market (see, e.g., Feniger et al., 2023).

Explanatory VariablesFootnote 1

Year of Arrival in Israel

This was represented by a series of dummy variables: 1990, 1991, 1992, 1993, and 1994. The wave of immigration from the FSU to Israel started in 1989, but the number of immigrants was relatively low. Therefore, we combined them with those who immigrated in 1990.

Control Variables

Parental Education

We distinguished between families in which at least one parent had an academic degree and families in which neither parent had such a qualification. This is our primary indicator of immigrant selectivity. Our preliminary analyses showed that using separate variables for the mother and father yielded similar results. Therefore, we chose the more parsimonious solution.

Number of Siblings in the Household

This was a continuous variable. In cases where information on the number of siblings was missing (about 15% of cases), we imputed the mean number of siblings and assigned the value of 1 in the dummy variable “Number of Sibling Missing.”

Age at Migration

This was represented by a series of dummy variables about each age at migration from 6 to 12. We decided to use dummy variables because we do not assume a linear relationship between this variable and the dependent variables. Additional analyses (not shown here) confirmed that the findings remain similar when using a continuous version of this variable. Age at migration is an important control because, theoretically, arriving at an older age may result in less successful educational integration in the host country (see Feniger et al., 2023 for FSU immigrants in Israel). It is important to note that for each year of migration in the sample, there is an entire range of ages at migration, and the distribution of age at migration for each year is quite similar. This means that the year of migration itself is not directly linked to the number of years an individual studied in the Israeli education system. Yet, we control for age at migration when testing the effect of the year of migration.

Gender

A dummy variable, where female = 1.

Data Analysis Strategy

We first present descriptive statistics for the dependent and independent variables, organized by year of migration—our primary explanatory variable. Next, we employ a multivariate analysis. We estimated linear regression models for the Hebrew score, a continuous measure. As a robustness check for the Hebrew score, we also estimated a linear model with a Heckman correction for selectivity for taking the test. The explanation for this model and its results are presented in Appendix 1. We estimated logistic regression models for eligibility for a matriculation diploma and attainment of an academic degree, which are dichotomous variables. For all three dependent variables, the first model includes only years and age at migration, while the second model includes these variables, parental education, and control for number of siblings and gender. Through the second model, we can determine if differences in parental education (i.e., changes in immigrant selectivity) can account for the patterns identified in the unadjusted first model.

Findings

To answer research questions 1 and 2, Table 1 presents descriptive statistics for our independent variables by year of migration. Table 2 gives descriptive statistics for our dependent variables by year of migration. While this study does not compare immigrants to non-immigrants, we added information on comparable non-immigrant Jewish cohorts in Israel in order to embed the findings regarding FSU immigrants in the broader context of Israeli society. In line with theoretical arguments on declining immigrant positive selectivity over time, Table 1 clearly shows that the average level of parental education considerably declined among FSU immigrants between 1990 and 1994. Furthermore, for the first two years, the average parental education among FSU immigrants was much higher than the average parental education in the general Jewish population. In contrast, in the last two years, parental education was lower than the average in the non-immigrant Jewish population. In the other independent variables, we did not observe a similar trend.

Table 1 Independent variables by year of immigration
Full size table
Table 2 Dependent variables by year of immigration
Full size table

Table 2 shows a parallel decline in education outcomes among children of FSU immigrants. The Hebrew score declined by about one-third of a standard deviation between the first and the last years of immigration in our sample. The chances of attaining a matriculation diploma and an academic degree were also lower for those who arrived in later years. Regarding the matriculation diploma, the percentage of FSU immigrants eligible for the diploma was higher than in the general Jewish population for the first two years and much lower than among this group in the last two years of immigration. Compared to non-immigrants, immigrants from all years of immigration had lower chances of attaining an academic degree, but the gap between immigrants and non-immigrants considerably grew over time. These findings should be interpreted against the general trend in Israeli society during the 1990s and early 2000s. These decades were characterized by a steady increase in both matriculation eligibility and higher education enrollment (Central Bureau of Statistics, 2010). This means that our estimations of the decrease in matriculation eligibility and attaining a higher education degree among FSU immigrants are conservative. Model 1 in Table 3 and model 1 in Table 4 reveal that the results regarding the decline in educational outcomes of immigrant children by year of migration remain the same even after controlling for age at migration.

Table 3 Linear regression models predicting scores in Hebrew matriculation exams
Full size table
Table 4 Coefficients from logistic regression models predicting eligibility for a matriculation diploma and an academic degree
Full size table

To examine to what extent the drop in mean educational achievements of children by year at migration can be explained by the decline in the mean education attainment of the parents (i.e., by a decline in immigrant selectivity), we conducted a series of multivariate analyses. Our first analysis used linear regression models to predict the Hebrew matriculation exam scores. Table 3 presents coefficients of the year of immigration (the reference year is 1991) using two models: (1) in the first, we controlled for the age at immigration; (2) in the second, we added parental education and also controls for number of siblings, and gender. Results from this analysis indicate that the trend we observed in the descriptive analysis is only partially explained by controlling for socio-demographic variables. After controlling for these variables, the coefficients of the years 1992, 1993, and 1994 were reduced by about 30–40% and remained statistically significant in the adjusted model. This means that while the change in parental education composition of the group of FSU immigrants (i.e., changes in selectivity) is an important explanation for the decrease in the Hebrew score, it cannot account for the entire drop in achievement. In Table 5 in the appendix, we present an additional analysis of the Hebrew score with a different modeling strategy. Instead of treating not taking the test as a zero score, we estimated a two-stage Heckman correction (Puhani, 2000) linear model that takes into account changes in the odds of taking the exam. As seen in the appendix, the results and the conclusion are similar to those of the models presented in Table 3.

Our next step of the analysis focused on attaining a matriculation diploma at the end of high school and an academic degree after that. For each dependent variable, we created two logistic models: in the first, we estimate the effect year of immigration controlling for age at migration; in the second, we added parental education and controls for gender and number of siblings. We found similar patterns for both educational outcomes. The odds of attaining a matriculation diploma and an academic degree decreased when the year of immigration was later than 1991. The models adjusting for family socio-demographics suggested that this drop in educational achievements was not fully explained by the declined positive selectivity of the immigrant group in terms of parental education. For both dependent variables, the coefficients of the years of immigration (after 1991) remained statistically significant after controlling for parental education, number of siblings, gender, and age at immigration.

To further explore the extent to which changes in the social characteristics of the FSU immigrant group accounted for the effect of the year of immigration on eligibility for a matriculation diploma and academic degree attainment, we calculated average marginal effects (AMEs) for the year of immigration coefficients for both models presented in Table 4. These AMEs are presented, with their corresponding confidence intervals, in Table 6 and Table 7 in the Appendix. Comparing the AMEs of the two models for each dependent variable enabled us to answer the third research question. In most cases, the decrease in the effect of year of immigration between the first and the second model was about 20–30%. The highest decrease in this effect, 45%, was for 1993 (compared to 1991) in the models predicting eligibility for a matriculation diploma.

We conclude, then, that changes in the social composition of the immigrant group, most notably in parental education, can account for a non-trivial decline in eligibility for a matriculation diploma and attainment of an academic degree among members of this group who immigrated to Israel as children. Still, in both cases, an important unexplained effect of the year of immigration remains that cannot be attributed to our control variables.

Conclusion

Immigrant selectivity has received considerable attention in recent years, but whether this selectivity changes over time remained relatively underexplored. As we showed in the literature review section of this paper, the few studies dealing with this question, some of them relatively old, have found that selectivity among immigrant groups tends to decline over time (e.g., Browning & Feindt, 1969; Massey et al., 1990; Rhoades, 1978). We developed this literature by focusing on the large wave of immigration from the FSU to Israel in the early 1990s and asked (1) whether the educational attainment of parents in this group (acquired in the FSU) decreased over time; (2) whether the educational attainment of the children of these immigrants (acquired in Israel) decreased over time; and (3) if the answer to the first two questions was positive, whether changes in parental education can account for the drop in children’s educational outcomes.

Our analysis of a large representative sample of FSU immigrants in Israel revealed a parallel decrease in parents’ educational attainment and children’s educational outcomes (Hebrew achievement at the end of high school, attaining a high school matriculation diploma, and attaining an academic degree) within a relatively short period of arrival (five years). The study thus corroborates older studies at different times and in other social contexts that have reported declining immigrant selectivity over time (Jones, 1998; Massey et al., 1990; Popkin, 1999; Rhoades, 1978). This means that developing a temporal perspective on immigrant selectivity is crucial, and researchers should not assume that the social characteristics of an immigrant group remain constant over time, even during a short period. This has important implications for comparing immigrant and non-immigrant groupscentral to immigrant selectivity studies. For example, in the Israeli case, immigrant students who arrived in 1990–1991 had higher chances of attaining a matriculation diploma than non-immigrants, while those who arrived only a couple of years later in 1993–1994 had much lower chances of achieving a matriculation diploma than their native-born counterparts. At the same time, the overall percentage of matriculation eligibility among non-migrants in Israel had increased (Central Bureau of Statistics, 2010).

We also asked whether immigrant selectivity among parents fully accounted for changes in educational outcomes over time among their children. We found that changes in immigrant selectivity accounted for only part of the decline in their children’s educational achievement. This finding has significant theoretical and policy implications, as it suggests that educational opportunities for immigrant children are influenced by a range of factors beyond immigrant selectivity. For instance, when large numbers of immigrants arrive in a receiving country, the local education system is expected to provide high-quality language instruction to accommodate the growing number of students. In the early 1990s, the Israeli education system experienced this challenging situation when tens of thousands of immigrant students from the FSU joined the education system each year (see Horowitz, 1999; Sever, 2004). While we were unable to explore the potential role of the education system in the decline of the educational achievements of FSU immigrants during the time period covered by our study, we suggest that saturation in the host education system may have had a negative effect. We hope our study will encourage further research into this topic in Israel and elsewhere.

This study also highlights the importance of taking a dynamic approach to understanding the characteristics of immigrant groups when developing social policies to integrate them into the host society. This is particularly crucial for groups with high levels of educational success, often referred to as “model minorities” (see, e.g., Poon et al., 2016). As demonstrated in this paper, the initial FSU immigrants to Israel were highly educated, and their children enjoyed high levels of success in the Israeli education system, but this changed over time, with subsequent immigrant students from the same group falling behind their native-born counterparts. It is essential, then, to recognize and adapt to these changes in order to effectively support the educational success of migrant students in the host country.