WORKING PAPER SERIES NO. 74 Emigrating Israeli Families: Who Goes Where? Characteristics of Families of Israelis Who Immigrated to Three Destinations: the United States; Central and Western Europe; and the former Soviet Republics and Eastern Europe Eilat Cohen-Kastro* ([email protected]) January 2013 * Israel Central Bureau of Statistics – Demography and Census Department 1 The Israel Central Bureau of Statistics (ICBS) encourages research based on ICBS data, such as the research published in this paper. These research papers are not official ICBS publications. Therefore, the opinions and conclusions expressed therein are those of the writers alone and do not necessarily reflect the opinions and conclusions of the ICBS. Israel Central Bureau of Statistics, 66 Kanfei Nesharim St., corner Bachi P.O.B. 34525, Jerusalem 91342 Tel. +972-2-6592666; Fax +972-2-6521340 www.cbs.gov.il [email protected] 2 Abstract This study examined the characteristics of families that emigrated from Israel to selected emigration destinations. The research was carried out through the use of official Israeli databases. A unique methodology was developed that allowed for the identification of the emigration destination country of Israeli families abroad. This was accomplished by using the listing of the country of birth that was registered for children born abroad to Israeli residents. This country of birth served as a proxy for the emigration destination country of the whole family. The population study consisted of 34,047 families of Israelis (133,502 individuals) to which a child was born abroad and whose birth was listed in Israel's Population Register, during the years 1996–2008. The research had two purposes: (a) producing a new database of Israeli families abroad—parents and their children up to age seventeen—during 1996– 2008, that included demographic and socio-economic characteristics as well as each family's emigration destination country; and (b) analysis of the emigrating families according to their selected emigration destination and according to various demographic and socioeconomic characteristics. Three emigration destinations were selected that represented distinct social and economic structures: (a) the United States (US), (b) Central and Western European countries, and (c) Eastern European countries and the former USSR republics. The study used two research methods: descriptive statistics to depict aggregated measures of the characteristics of the families that had emigrated by their emigration destination, and binary logistic regression to determine the most influential characteristics. Variables that could have an effect on the decision to emigrate and its destination were chosen to be used in both methods: religion, country of origin, immigration history, education, and occupation. The results showed that the Israeli families that emigrated to the US were characterized, on average, by a very high level of skills. Families that emigrated to Central and Western European countries were not characterized by a high level of skills and families who emigrated to Eastern European countries or the former USSR republics were characterized by relatively lower levels of skills. Moreover, families whose origin was Eastern European countries or the former USSR republics tended not to return to their country of origin but rather to emigrate further to a more rewarding destination. It was concluded that the Israeli families that emigrated to the US were characterized by a high level of skills. They constituted the majority of the emigrating families as well as the majority of the highly skilled ones, in comparison to families that emigrated to the other 3 destination that were studied. The new database of Israeli families abroad that was constructed for this study enables further research with regard to emigration from Israel and study of the family as a migration unit. Key words: immigration, emigration, destination of immigration, Israeli families abroad, religion, country of origin, immigration history, skill level, education, occupation. 4 1. Introduction More people today live outside their countries than ever before in history, and the number of people who move across international borders is expected to increase even more in the future (United Nations, 2009b). In recent decades, international migration has become a topic high on the international agenda, because the implications of migration are relevant to many "hot" topics, such as low fertility rates over time and aging of populations, unemployment, "brain drain" and "brain gain", transfer of migrant remittances, human rights, social integration, fear of foreigners, human trafficking, and national security. The implications of international migration on these and other issues influence the populations of both the countries of origin and the destination countries of migrants who move across international borders. According to the United Nations, international migration is defined as a change a person's place of residence from one country to another. It includes people who leave one country to live in another (United Nations, 2009a). From a demographic perspective, because this movement affects the population of both countries, it must be recorded twice, once as immigration to the destination country and once as emigration from the country of origin. However, there is a certain amount of asymmetry between the accessibility of data on emigration and data on immigration, due to two major reasons: (1) Exits tend to be documented less well than entrances, because most governments hesitate or are unable to track closely emigration of people from their territory, especially if those exiting are their citizens and not citizens of a foreign country. (2) From a statistical perspective, it is difficult to count people who left a country, because they are not present. This comes into play in measurement of both movements and stock. Censuses and sample surveys, which collect information on the population of residents, inherently involve problems in counting people who are not present, especially when the members of the household no longer live in the country of origin (United Nations, 2009a). 5 Israel, in contrast to most of the countries which at certain times limited the incoming of new immigrants, has, since the establishment of the State, had a policy that has continuously encouraged unlimited Jewish immigration. In accordance with the goal of increasing the Jewish population, Israel attempted to discourage (Jewish) emigrants, generally by employing ideological and moral pressure. Since Israel's establishment, the phenomenon of emigration has attracted public attention. Emigration of Jews from the country is considered as weakening it and the Zionist ideology behind its establishment, because the State was founded based on the ideals of "the ingathering of the exiles" and establishing a national home for the Jews in Israel. In Israel, neutral demographic concepts such as "immigration" and "emigration" were exchanged for value-laden concepts with positive connotations for immigration (aliya or "ascending") and negative stigmatization for emigrating (yerida or "descending"). Therefore, it is not surprising that in Israel, emigration is considered a social problem, and has engendered much literature. Until the 1980s, this literature focused mostly on the "severity of the problem" and attempts to understand why Israeli Jews, especially natives of Israel, preferred to leave the country. As the country matured and Israeli researchers adopted more scientific standards to study this demographic phenomenon, an increasingly large body of literature developed in which it was concluded that the rising emigration rate and the number of Israelis living abroad were not unusual in their size, given the unique circumstances in Israel: a large percentage of residents who were born abroad, no opportunity of short-term migration to neighbouring countries, high rates of return migration, and ongoing conflict with the Arab world (Cohen, 2009; Eisenbach, 1989; Michaeli, Epstein, & Kheimets, 2007; Sikron, 2004). The Israel Central Bureau of Statistics (ICBS) currently does not study emigrants from Israel in terms of the various emigration destinations. The focus in the statistical data that are published is on the total movement of departing and returning Israelis at the border, and an estimate of the stock of Israelis who reside abroad and are "suspected" of being (i.e., might possibly be) migrants who settled in another country, with analysis by some basic 6 demographic characteristics such as age, gender, marital status, and origin. The ICBS defines Israelis staying abroad as Israeli residents who departed abroad and stayed there more than 365 days continuously, and who stayed in Israel for at least 90 consecutive days before their departure (Hleihel and Ben-Moshe, 2002). Characterizing emigrants leaving Israel is challenging for various reasons. An Israeli resident is not required to declare the purpose of a trip (e.g., emigrating, touring, temporary work) when leaving the country. There is no obligation to declare that one is leaving Israel for the purpose of settling in another country. Furthermore, often residents do not know when leaving Israel whether they will return to live there. Often, the emigrants continue to maintain ties to Israel: they visit Israel often and sometimes even still own an apartment there. They have dual ties, to both Israel and to the destination country. In addition, in cases in which the emigrants visit Israel often, it is difficult to determine whether their "centre of life" is in the country of origin or the destination country. Also, more than a small number of Israeli residents depart abroad and return to Israel after a prolonged stay abroad. Thus, one aspect of the problem is that it is difficult to estimate the overall number of emigrants, and another aspect is that currently there is no way to study their characteristics with respect to emigration destinations based on official Israeli data. The first aspect is handled adequately, within the existing limitations, by the estimates published by the ICBS. The current research deals with the problems of the second aspect described above: investigating characteristics of emigrants from Israel with regard to their destinations. The idea of developing a methodology enabling the identification of the destination country of Israeli families abroad was conceived during investigation of the issue of emigration from Israel. Briefly, the methodology uses the country of birth of a child born abroad to Israeli residents to act as a proxy for the destination country of the entire family (detailed descriptions can be found in the Method section and in Appendix A). Based on use of this methodology and on data existing at the ICBS, it was decided that the research population would be families of Israelis abroad to which a child was born and whose birth 7 was recorded in the Population Register, during 1996–2008. To the best of my knowledge, this is the first usage of a methodology that provides information about the destination country of emigrating resident Israel families, from an Israeli data source. Once it was possible to identify the families' destination countries, the research goals of a study of characteristics of families of Israelis abroad were defined as follows: (a) the creation of a new database of families of Israelis abroad (parents and their children up to age 17), during 1996–2008, including demographic and socio-economic characteristics as well as the destination country of each family in the database, and (b) analysis of families that emigrated, by selected major destinations. Attaining the first research goal, which was the creation of a new database as a result of development of the methodology, conferred several advantages compared to foreign sources of data (from foreign countries and international organizations), which were used in previous research on emigration of Israelis. First, the OECD's statistical database, the Database on Immigrants in OECD Countries (DIOC), enables identification of only native Israelis aged 15 and over who emigrated to an OECD country. The new database created in the current research also includes children aged 0–14. Second, the DIOC database presents aggregated statistics that do not allow breakdown by families and their differing compositions. The new database allows breaking down by families according to any composition needed (including families in which one spouse or both are not native Israelis). Third, the DIOC database does not include data on all of the destination countries, even if they are OECD members. For example, because Germany and Netherlands do not collect data on the birth countries of their residents, one cannot identify natives of Israel who emigrated to these countries using this database. The new database includes all of the countries, as they appear in the variable "country of birth" recorded by the Ministry of Interior and included in the Israel's Population Register. Fourth, when relying on foreign sources of information, such as national censuses or data from foreign national bureaus of statistics, one must rely on definitions, legislation, types of data collection systems, and frequency of 8 data collection based on each country's discretion. In the new database, determination of the research population is based on Israel's Population Register whose definitions and updating methods are known, and are accepted by both the authorities and the researchers in Israel. This enables research on a population of Israelis or former Israelis using standards that are accepted in and appropriate for Israel. Fifth, in foreign national censuses, one can usually locate the population of Israelis or persons who emigrated from Israel only during the past five years, and even then, one cannot distinguish directly between the different types of population groups in Israel. The new database enables characterization of all persons who departed Israel who are included in the research population and are listed in the Population Register, even if they immigrated in the past to Israel from another country and left Israel after some time period for another destination. In addition, the data from the Population Register allows one to distinguish among various population groups in the research population (for example, to characterize according to religion, which enables distinguishing between Jews and Arabs, which could have different demographic and socio-economic characteristics). Sixth, inclusion in some foreign databases, such as that of the United States immigration department, depends on the immigrant's legal status. If a person's stay in the destination country does not involve a statutory residential status, the immigrant will not be included in the database. Inclusion in this new database is not dependent on the statutory status of the immigrant in the destination country, rather, only on whether the immigrant is included in Israel's Population Register and whether he or she has child(ren) who were born outside of Israel, the record(s) for whom are also included in the Register with a status that indicates that the parents live or lived outside of Israel. Attainment of the second research goal, namely, characterization of families of emigrants according to their destination countries, increases our knowledge about the emigrant population from Israel in general, and in particular, according to selected destination countries of the emigrants. In addition, use of the family as the investigation unit enables research on various levels: one can investigate characteristics of each individual in the 9 family separately and the variance in the characteristics between that person and other individuals, and concurrently investigate characteristics of the family as one unit and the variance in characteristics between that family and other families. There are a number of advantages to characterizing the emigrating families using the new database: First, focus on emigration as a family unit. Investigation of an emigrant in the family context in which he or she lives is necessary in order to understand the decisions made about emigration within the family unit, and which are influenced by the family structure and by the characteristics of the family as a whole. Second, family emigration has characteristics of its own, which do not necessarily match the characteristics of lone emigrants or couples without children. It is my opinion that there is important potential that has not been utilized up to now in focusing on the family aspect of emigration from Israel, which involves characteristics that are common to the entire family unit, as well as characteristics that are unique to individuals within families. Third, examination of the demographic and socio-economic characteristics of Israeli families by destination country of emigration (specific or grouped) enables pinpointing differences in family characteristics by emigration to different destinations, thus increasing the possibility of research that takes into account economic and social characteristics of the structure of the destination country among the various reasons for emigration, leading to a fuller picture of the phenomenon of emigration from Israel. In accordance with the theoretical framework detailed below, three emigration destinations were selected whose economic and social characteristics differ from each other. Characteristics of the families which emigrated to these countries were examined separately: (a) the United States, (b) countries of Central and Western Europe, and (c) the republics of the former USSR and Eastern Europe. The reasons for selection of these destinations are explained in the research hypotheses, and a list of the countries included in each destination appears in the Method section. We continue with a literature review, which presents relevant approaches from the theoretical frameworks that are used in research on international migration, followed by 11 major topics from the theoretical frameworks dealing with research on emigration from Israel. The research hypotheses, which were formulated based on the theoretical frameworks, are presented. Then, the methodology that was used to investigate the hypotheses is presented. The following chapter presents the results of the research, according to the variables that were selected to characterize the families at each destination. Finally, there is a discussion and summary of the characteristics of the families of Israelis who chose to emigrate to each destination, and directions for future research are presented. 1.1 Theoretical framework Large migration movements, which began after 1945 and have expanded substantially since 1980, are taking place throughout the world, whereas previous large migrations were mainly trans-Atlantic. Against this background, Castles and Miller (2010) referred to the contemporary era as the "the age of migration". One of the reasons they mention for this migration is increased mobility, which has been facilitated by political and cultural changes. In addition, they mention the development of new transportation and communication technologies, which have made it easier to plan and implement migration. Globalization is another factor that has promoted and expanded international migration. Migration movements can take many forms: People can migrate as manual labourers; as specialists with high-level skills; as entrepreneurs; as refugees; or as family members of previous migrants. The migrant's status in the country of origin also plays an important role: Destination countries vie with each other to attract highly skilled immigrants through legislation providing privileged entry and residence to this population. In contrast, manual labourers and refugees often experience exclusion and discrimination. These processes have been accompanied by the development of new types of migration: retirement migration; migration for upward mobility and a better (or different) life style; and repeated or circular migration. The distinction between migration and tourism has become blurred, as many people travel as tourists in order to explore potential migration destinations. 11 The present "age of migration" reflects the basic socio-economic structure of almost all industrialized countries, most of which have become diverse and multi-cultural as a result of this process. Even though there is no single theory that explains international migration, several theories have attempted to provide explanations about the primary motives for migration; and other theories have attempted to explain why the migration has continued, expanded, and then developed. The explanations underlying the motives for migration are mainly economic, and derive from theories about the development and conservation of the capitalist economy. The first theory that aimed to explain motives for migration was the neoclassical theory of economics: macro and micro theory. According to this theory, international migration stems from geographic differences in the supply and demand for labour. Differences in wages cause workers from countries with low wages to emigrate to countries with high wages. The theory posits that labour markets are the main mechanism that motivates migration (the macro-approach). Accordingly, international migration derives from differences in wages and employment across countries (without assuming that there is full employment), as well as from a rational decision-making process. Because "rational actors" make individual choices, the reasons for migration are based on cost-benefit calculations, which lead to the expectation for positive net returns (usually financial compensation) for migration. That is, taking their skills into consideration, potential migrants estimate the advantages and disadvantages of moving to different countries, and choose to emigrate to the country that provides the highest returns in the foreseeable future (the micro-approach)(Massey et al., 1993). According to this approach, the decision to migrate is based on two major variables: (1) differences or gaps in earnings and other forms of returns in the destination country versus the country of origin; (2) the feasibility of finding employment in the destination country. In other words, the decision to emigrate depends on the income that a person can expect to earn in the destination country, in relation to the income that can be expected in the country of origin. According to this model, migration is defined as the economic 12 investment in human productivity. Like all forms of investment, it has costs as well as benefits. Rational actors anticipate these costs and benefits when they decide whether and where to emigrate. Studies founded on this theory have examined the effect of several individual characteristics on migration, such as age, education, marital status, work experience, and unemployment (see, e.g., Massey, 1990; Todaro, 1969; Todaro & Maruszko, 1987). Another approach that focuses on the reasons for migration is self-selection theory, which attempts to explain the extent and composition of streams of international migration (Borjas, 1994). According to this approach, there is a large pool of people who believe that there are better economic opportunities elsewhere, and who are willing to bear the costs of pursuing these opportunities. Both of these conditions indicate that the pool of potential migrants is not selected randomly from the population in the countries of origin, nor is it randomly spread across potential countries of destination. Rather, the pool of migrants in every destination country consists of a sub-sample of people with skills that give them better opportunities in another destination country than they have in their country of origin. In addition, the costs of their migration are low enough to make the transition profitable. In the context of the United States, which is the country with the largest immigrant population in the world (UN, 2009a; many studies cited in the literature have used data produced from the US Census), two conditions are necessary for positive selection of potential immigrants: (1) a strong positive correlation between the immigrant's expected earnings in the country of origin and the expected earnings of the same immigrant in the United States; (2) a more unequal distribution of income in the United States than in the country of origin. Consistent with this argument, examination of empirical data reveals that the United States attracts immigrants with above-average skills, and that the characteristics of the country of origin explain a large share of the differences in the economic status of groups of immigrants to the US. This is because the skills acquired in developed economies can be transferred more easily to the American labour market. Thus, there is a strong positive correspondence between the 13 migrant's earnings in the US and the level of economic development in the country of origin. More specifically, migrants whose earnings in the US are relatively high in comparison with their measurable skills arrive from countries with high levels of GNP, low levels of income inequality, and politically competitive systems (Borjas, 1987, 1988, 1994; Jasso & Rosenzweig, 1990). In contrast to these approaches, which examine migration based on the individual as a unit of analysis and argue that the goal of migration is to maximize returns, other researchers have argued that households or families are the main agents that make decisions, and that migration should be perceived as part of broader strategies for socio-economic preservation and improvement. In accordance with these approaches, empirical models of migration behaviour should be examined on the basis of households or families as the unit of analysis (Massey, 1990). This is a broader perspective, which considers reasons for migration that go beyond the rational decision of the individual, and includes the family as an integral part of the decision-making process about migration. As such, the New Economics of Labour Migration theory shifts the focus of migration from individual independence to mutual interdependence. Stark and Bloom (1985) defined migration as a "calculated strategy" in mutual relationships, and not as an isolated activity aimed at unlimited maximization of possibilities. According to this approach, the family is not a divided or separate entity, in which the young members seek independence and leave in an attempt to separate themselves from traditional family ties and obligations. Rather, the family is viewed as a social and economic unit, which exists in an unstable environment and chooses to disperse its financial risks as an extended unit across various sources of income. This is done by financing emigration to a more developed country for one family member, who will enter an employment sector which provides an income that correlates negatively with the same sector in the country of origin. The emigrating family members, who will earn more than they did in their country of origin, transfer a substantial share of their income back to the family members who stayed in the country of origin. In that way, both parties benefit from the 14 emigration because the exchange of obligations to share remittance provides mutual security. The shared obligations might explain the characteristics of the immigrants' financial performance. The migrants often outperform the native-born in the destination country. In addition, strong reliance on the "capital of social networks and family members" is also an important characteristic of the behaviour patterns of migrants (Stark & Bloom, 1985). Notably, in general, there is a lack of comprehensive research literature on migration of families, which is the focus of this paper. There is still much room to develop approaches which will integrate various interests, skills, and characteristics that can exist within the same family and have a differential impact on migration patterns. In my view, many migrant families that have high skills show a greater tendency to migrate as a family as opposed to migrating as individuals because they have the appropriate resources and legal status. From a structural perspective, world systems theory, which is based on the well-known research conducted by Wallerstein (in Massey et al., 1993), associates the sources of international migration with the world market structure that has developed and expanded since the 16th century. According to this theory, the penetration of capitalist economic relationships into peripheral noncapitalist societies creates a mobile population that is prone to migrate abroad. In the global system, international migration is a by-product of disruptions and dislocations that occur in the process of capitalist development. The more capitalism has spread outside of the core of Western Europe, North America, Oceania, and Japan, the more the population in the rest of the world has integrated into the global market economy. When land, raw materials, and labour in peripheral areas are affected and controlled by market processes, there are inevitably streams of migration abroad. Wealthy core countries have sent links of goods and capital to less developed countries through international structured pathways; and in the opposite direction, labour forces have been sent to the core countries through the same pathways (Massey et al., 1993). As mentioned, globalization has linked the economies of different countries, and has generated more mobility of people, capital, and information. In the Israeli context, the globalization process has gained 15 momentum since the 1990s, and the Israeli economy has become more international in terms of the entry of international corporations on the one hand and in terms of Israeli corporations expanding into external markets on the other hand. This has brought into the Israeli economy international corporations, mostly in the fields of communications and hightech, which in turn facilitates or migration (including long-term) between countries (Lahav & Arian, 1999). The theories that explain the continuity and perpetuation of migration include network theory, which argues that networks of migrants constitute a system of interpersonal relationships that link migrants, former migrants, and non-migrants in the country of origin and destination country through kinship, friendship, and shared community origin. These social networks play a functional role in constructing decisions to migrate, because the existence of friends, relatives and other members of one's personal community in the destination country substantially increase the probability of migrating there. These social networks also lower the costs of moving, and increase the expected net returns of migration as a result. As such, these networks constitute social capital that migrants can rely on in order to gain access to employment in a foreign country (Massey, 1990; Massey et al., 1993). Despite the differences in the theoretical approaches described above, it is important to note that international migration takes place on different levels simultaneously. On one level, it is driven by individual motives (personal considerations such as costs and benefits, maximizing potential returns, and mutual dependence of families). Concomitantly, migration is a result of social and economic structures (e.g., supply and demand in the labour market, wage gaps in different countries, and globalization processes). In addition, there are push and pull factors for migration, such as social networks that are structured in terms of their range and scope, as well as individual factors such as personal contacts and origin. However, the existence of international migration is not necessarily an outcome of all these factors together. It is possible that in certain cases, one type of factor will have a significant impact on migration from country to another whereas other factors will have less of an impact. 16 Besides the factors influencing migration presented in the above theories, which mainly reflect an economic perspective, several other factors affect streams of migration between countries throughout the world. People can seek to migrate for personal reasons, e.g., because they are refugees or political asylum seekers who have been persecuted in their countries of origin; or because they seek to become reunited with their families. Political instability and violation of human rights in certain parts of the world are a push factor that has created forced movement of populations, usually on a large scale. The United Nations High Commissioner for Refugees (UNHCR, 2011) reported that there were 10.5 million refugees throughout the world in 2010. With regard to Israel, the UNHCR reported that 1,301 refugees and 799 political asylum seekers left Israel over the past 10 years. Another factor that influences migration is ideological background. Although the large waves of immigration from Europe to the United States, Australia, Canada, and other countries were not motivated by ideological considerations, the situation was different in Israel. In the former case, leaving one's country of origin or returning to one's country of origin due to adjustment problems in the destination country was not associated with negative characteristics of the immigrants themselves or of the countries of migration. Even though emigration from Israel (yerida) has occurred since the beginning of the Zionist movement, the attitude toward people who leave the country has not been tolerant. Moreover, people do not usually leave Israel for ideological reasons (e.g., disappointment with Zionism or a sense of alienation from the country). Rather, in most cases people leave due to their personal circumstances and out of a desire to improve their economic situation. Nonetheless, in the 1950s, 1960s, and 1970s, emigration from Israel was still considered to be a loss of one's Israeli identity. In subsequent decades, the stigma associated with yerida declined; the public has shown a tendency to accept emigration from Israel as more legitimate, and has expressed more willingness to understand and to consider the circumstances underlying the decision (Michaeli, Epstein, & Kheimets, 2007). Whereas most people do not migrate for ideological reasons, there are those who also seek to leave their country of origin because 17 their personal ideology conflicts with the ideology of that country. In a paper dealing with emigration from Israel to the United Kingdom, Gold and Hart (2009) argued that despite the negative stigma associated with the concept of yerida in Israel, recent political and economic crises have made Israel a less desirable place to live and have caused some residents to seek better living conditions abroad. The authors indicate that toward the end of the first decade of the 21st century, political and economic crises became a push factor for emigrants to leave Israel, and, at the same time, the number of emigrants who have returned to Israel is lower than it was in the mid-1990s. Attitudes toward emigration from Israel today have been affected by the more tolerant attitudes toward yerida that developed during the 1990s; in contrast to the past, emigrants today can criticize Israel publicly and justify their decision to leave the country. In a similar vein, Cohen (2009) argued that the security and economic situation of Israel explains a large share of the difference in the number of Israelis who left versus the number of Israelis who returned to the country. The highest rates of emigration and the lowest rates of returning Israelis were in 2001-2002. These were the peak years of the Second Intifada, when Israel faced the threat of constant terror attacks and the onset of economic crisis. At the same time, there was a period of a major influx of immigrants who arrived in Israel, some of whom stayed for a short time and then moved on to other destinations. This is a common characteristic of migration known as transit migration, and is another factor that caused the increase in the number of emigrants from Israel at that time. In addition, there is "circular migration", which refers to repetitive movement of migrants for continuous periods between the origin and destination countries, usually for purposes of work. Examples include academic professionals, scientists, and physicians who are on sabbatical abroad, and return to the origin country when the sabbatical is over (Cohen, 2008). According to these studies, some people who emigrate from Israel apparently have ideological motives, but this does not typify most of the emigrants, who leave for economic and employment reasons. 18 According to UN statistics, migration to developed countries remained high between 1990 and 2005 (United Nations, 2009b). The largest increase in the number of migrants was to North America, where 17 million immigrants were absorbed during that period, and to Europe (15 million immigrants). A large share of the migrants who arrived in those countries moved for the purpose of reuniting with family members, although the proportion of migrants who were categorized as highly skilled increased. In this case, another factor that affected the direction and nature of international migration was the absorption policy of the destination country. The governments of many destination countries took measures to promote the entry of migrants who are in demand, particularly educated migrants and lowlevel temporary workers. In 2005, a growing number of countries had a policy for promoting the immigration of highly skilled workers. These countries implement a selective immigration policy, which gives priority to migrants with skills that are needed in the science and technology sectors (United Nations, 2009b). It has been argued that the rapid growth of the information technology sector and the decline in fertility rates in Western countries (e.g., Canada, Germany, and the United Kingdom) caused these countries to adopt a migration policy that encourages absorption of highly skilled workers (Gould & Moav, 2007). According to UN statistics, approximately 30 countries (including 17 developed countries) have policies or programs to promote the entry of highly skilled workers. For example, this type of immigration policy is implemented in North America, where, in Canada, candidates for permanent residency are selected on the basis of a score system, in which higher scores are allocated to people with diplomas or degrees that are recognized by the Canadian authorities. Similarly, in the United States, there are several categories of visas that allow temporary entry to highly skilled workers. In Europe, several countries (France, Germany, the United Kingdom, the Czech Republic, Netherlands, and Sweden) as well as the European Union have eased the restrictions on entry and migration for highly skilled workers. These countries have recently revised their approach to migration by establishing specific pathways for highly skilled immigrants. Concomitantly, developed countries have sought to restrict and control the entry of low-skilled migrants. Therefore, developed 19 countries that absorb immigrants have been adopting a dual approach, which encourages the entry of highly skilled immigrants on the one hand and restricts the entry of low-skilled immigrants on the other hand (Castles & Miller, 2009; Skeldon, 2009; United Nations, 2009b). It is clear that the immigration policy of the destination countries has distinguished between different types of immigrants. However, where should highly skilled people with no personal, ideological, or other restrictions seek to migrate, given that most countries would welcome the highly skilled migrants? A partial explanation for this, which relates only to the families included in the research group, can be found in the results presented in this paper. The findings indicate that people who seek to migrate face various constraints and barriers that can delay or prevent their migration to certain destination countries. The personal or ideological motives for migration during certain periods might lead them toward a certain destination country where entry is restricted to them due to immigrant absorption policies or other restrictive conditions in that country. In those cases, the person will either choose not to emigrate at all, or will emigrate to another country. In any case, people who seek to migrate make certain choices, and this paper will examine the characteristics of people who choose to migrate for an extended period, and sometimes permanently, to one of three destinations. In so doing, the paper aims to examine the characteristics of families that were included in the research population and emigrated to a certain destination country. Before presenting the specific research hypotheses, methods, and results, the following section provides an overview of emigration from Israel. 1.2 Emigration of Israelis, and Research Hypotheses Researchers have tried to characterize emigration of Israelis and estimate its impact on Israeli society in various areas. For example, Gould and Moav (2007) found that a high percentage of Israeli emigrants are highly educated, and they referred to emigration of educated Israelis from Israel as a "brain drain". They also found a positive correlation between the tendency to emigrate from Israel and level of education: Israeli emigrants do not represent the whole population because the percentage of emigrants with high levels of 21 education is disproportionate. The United States is a preferred country of destination for educated Israelis, among other reasons because it is a centre for studies, teaching, and research at well-known, prestigious universities and scientific research institutes. It has been estimated that thousands of researchers, university faculty members, and academic scholars leave Israel every year, mainly for the United States. This estimate was presented in an article summarizing a panel on the brain drain from Israel, which was published in the Calcalist economic newspaper (Sikuler & Kurz, 2010). In that panel, Israeli researchers who returned from the United States shared their experiences. Some of them indicated that the decision to move abroad derived from a desire for academic advancement, and mentioned that there is an unwritten assumption that in Israel, if you want to pursue an academic career you must have studied abroad at some point. According to those participants in the panel, there are many academic scholars who do not reject from Israel, but who travel abroad in order to advance their career and do not succeed in returning to Israel for several reasons: the gap between Israel and countries abroad in terms of working conditions, professional and career opportunities, and personal considerations. The article quotes a survey conducted by the Israeli Ministry of Industry, Trade and Labor which revealed that only 2% of the researchers who went abroad in recent years feel they can return to an appropriate job in Israel; 90% reported that they had to compromise on their income and on opportunities for professional challenge. There are those who believe that emigration of Israeli scholars to the United States is desirable because academic scholarship is an international field, so in order to expand knowledge, universities need to send faculty members abroad and receive them when they return. In order to maintain an academic system with an international reputation, it is necessary to adopt this model. Throughout the world, there are numerous scientists who study in the United States, and not all of them return to their countries of origin. The researchers interviewed on the Calcalist panel view this as a normal and acceptable phenomenon. However, to Israel's credit, they also argue that the level of the country's academic institutions is high and that many scholars who study abroad still consider Israel to be their home (Sikuler & Kurz, 2010). 21 The relatively large percentage of emigrants from Israel with high levels of education is not unusual. The theoretical and empirical literature cited above indicates that, consistent with neoclassical theory and the self-selection theory, if the return for higher skills is greater in the destination country than in the country of origin, those who choose to emigrate will have higher levels of skills and income. In this respect, Israel is not different from other countries. That is, emigrants from Israel with high skills, statutory status in the destination country, as well as academic degrees and skills that are in demand, show a greater tendency to emigrate while maintaining ties with their country of origin. In contrast, low-skilled migrants who do not have statutory status as citizens outside of Israel, are more likely to find migration to be much more expensive, problematic, and temporary (Gold & Hart, 2009). Emigration of academic scholars from Israel is a global phenomenon that can be expected to increase over time. The catalysts for this process include the expansion of the IT sector, which has made job-seeking and migration to a foreign country easier than ever, as well as the rapid growth of the high-tech sector and the decline in the fertility rates in Western countries (e.g., the United States, Canada, Germany, and the United Kingdom), which has caused these countries to adopt a migration policy that encourages the absorption of highly skilled workers. Research conducted among Israeli-born residents of the United States (based on US Population Census data) has revealed that their education level is relatively high. The largest difference between Israeli-born residents of the United States and the comparison group of Israeli residents related to the level of academic education (MA degree and higher), which was higher for those residing in the US. Considering the high education level of Israelis, it is not surprising that their performance in the American labour market was found to be higher than the performance of American residents born in the US. Both male and female Israelis who immigrated to the US were found to have higher-level socioeconomic characteristics than white, non-Hispanic Americans born in the US (Cohen, 2008; Cohen & Haberfeld, 2003, 2007; Eisenbach, 1989; Michaeli, Epstein, & Kheimets, 2007; Toren, 1980). Cohen (2009) examined patterns of selection among Israeli-born migrants aged 15 and over in OECD countries. The findings of that study showed that the self- 22 selection of emigrants from Israel, which is measured on the basis of education, employment, and age, was most positive in English speaking countries, and especially in the US where the return on skills was highest. In contrast, Israeli emigrants with the lowest skills chose to migrate to Scandinavian countries where the labour markets are relatively inflexible and the return on their skills tends to be lowest. Self-selection of skill levels in other European countries tended to be moderate. However, developing economies, which lack the regulation mechanisms and equality, such as in Eastern Europe, appear to attract few Israelis, and much less so, highly skilled Israelis. These findings are consistent with the theories of self-selection in migration, which predict that highly skilled migrants will choose destinations where the returns for their skills will be highest. Regarding migration of highly skilled Israelis to the United States, a combination of motives was found, which have also been found in the theoretical literature. First, according to the neoclassical theory of economics, these migrants consider the United States as the destination where the returns for their skills will be higher than in other destination countries. According to the positive self-selection theory, highly skilled Israelis who choose to migrate to certain destination countries are not selected randomly from a pool of all Israelis. Rather, there are two conditions that lead them to prefer certain countries, and particularly the United States, as their destination: (1) a strong correlation between the income that these Israelis can expect to earn in Israel and the income they can expect to earn in the United States (in both cases, above-average income); (2) economic inequality is more prevalent in the United States. Since Israel has become part of the global economy, there are highly skilled Israeli migrants who are employed at institutions of higher education abroad, or who are employed at business and global corporations. These migrants have jobs that require them to leave Israel for an extended period with their family members (also referred to as relocation). Often, they move to the United States, where high-level technological, scientific, and business knowledge are concentrated. In the long run, this type of relocation can yield two different outcomes. First, it has the potential to lead to return migration to the country of 23 origin after their academic program or work assignment abroad finishes. This experience can improve their economic status in Israel, or it can lead to circular migration, which is characterized by intermittent periods of residence in two or more countries. Second, staying abroad for an extended period can create permanent migration to the destination country, with the assistance of social networks that the migrants establish during their stay abroad. In keeping with the new economics model of migration, all of the migration possibilities considered in this paper are based on the family unit. In this context, the theoretical and empirical literature highlights the importance of the nuclear family as the core unit for decisions about migration. Therefore the nuclear family (parents and children) was the basic research unit in this study, and the hypotheses are tested on the basis of a combination of the characteristics of each parent. The hypotheses of the current study, which maintain that migrants to the United States will have the highest skill levels, were derived from assumptions proposed in previous theories and in empirical research. The United States has been a major destination where people can maximize their returns, according to neoclassical theory and the self-selection theory. The United States is also a primary location for international business, and technological and academic research, all of which comprise a major aspect of the global economy. Concomitantly, and even as a result of this predominance, social networks expand with the growing migration of highly skilled Israelis to that country, and serve as a resource for other Israelis who migrate there. Based on all of the above, the United States was the first destination selected for this study. (Canada was not included because the level of returns and economic structure there differ from those in the US.) The second group of destination countries, in Central and Western Europe, are also characterized by a relative high level of economic development, but the structure of the labour market does not provide a level of returns that is as high as the level provided in the US. The third group of destination countries, in the former USSR and Eastern Europe, was chosen in order to examine whether the large number of immigrants who arrived in Israel from there will choose to return there, 24 and in order to examine the characteristics of migrants to countries that have a less developed economic structure than Israel in principle. The social networks mentioned in the literature review also encourage migration to the destinations that were chosen. There are migrants who arrived in Israel in the past from Central or Western Europe, and even more so from the former USSR and Eastern Europe, who could show a tendency to move back to their country of origin because they were dissatisfied with Israel and wanted to return to the supportive social and family networks that they had left behind. Theories that place the family at the centre of migration are consistent with the framework of the current study, which focused on the family as the main unit of research, and related to variables shared by the family as a whole as well as to variables that are specific to each individual family member. In accordance with the above, three research hypotheses were formulated: (1) Emigration to the United States: In accordance with the theories of neoclassical economics and self-selection, which maintain that emigrants will choose to emigrate to a country where their skills will lead to maximum returns, the research hypothesis stated that families that emigrated to the US will tend to be characterized by a very high level of skills: a level that will raise to a maximum the returns expected from the move. In addition, US natives who immigrated to Israel in the past and then emigrated from it will tend to return to their country of origin with their family members, for two reasons: (a) social and family networks that exist for them there, and (b) economic returns at a higher level than in Israel. In contrast, the hypothesis stated that natives of Israel (who do not have a supportive family network outside of Israel) with a high level of skills will tend to emigrate to the US, with the thought that once one decides to emigrate, one should emigrate to the destination with the best returns possible. (2) Emigration to Central and Western Europe: These countries are characterized by developed economies that basically can provide returns at a higher level than in 25 Israel, for emigrants with a high level of skills. Therefore, in accordance with the theories of neoclassical economics and self-selection, here the research hypothesis stated that the skill level of families that emigrate to these countries will be relatively high. Natives of these countries will tend to return to their country of origin in Europe due to the social and family networks that exist for them there, and also because of the shared regulations of free passage among countries of the European Union, which allow EU citizens to reside and work in any of the EU countries, which might raise the returns expected from the move. (3) Emigration to the former USSR and Eastern Europe: These countries are characterized by less remunerative economies than the first two destinations, and sometimes by economic, social, and political instability. Therefore, in accordance with the self-selection theory, the research hypothesis stated that families with a relatively low level of skills, especially those that originated in these countries, will tend to emigrate to a republic of the former USSR or Eastern Europe, where the expected returns are relatively low and which therefore do not attract emigrants with a high level of skills. There are two major reasons for emigration to these destinations: (a) the existence of a supportive social and family network in the country of origin, which reduces expenses and eases the move and integration, and (b) return migration to the country of origin on the part of persons who immigrated to Israel in the past and who did not acclimate well. The hypothesis was that many of them will be non-Jews whose connection to Israel is likely to be weaker. 26 2. Method As mentioned, the purpose of this research was to examine the characteristics of the group of Israeli families abroad in three selected emigration destination, during the years 1996– 2008. The research was enabled by the development of a methodology that permitted the creation of a new database that included the destination countries of emigrating families, including characteristics of the families and of the individuals in them. After investigation of various possible methods to research the research hypotheses, two statistical research methods were chosen: (a) descriptive statistics, to present aggregate measures of the characteristics of the emigrating families, by the various destinations, on order to obtain an overview of the status of the phenomenon using diagrams and statistical measures of various types, with the goal of the descriptive statistics being to describe the data, and (b) binary logistic regression (using the stepwise method, and using selected variables that were important to characterization of the phenomenon). This type of regression was chosen because the event of emigration of a family to a particular destination constitutes a dependent dichotomous variable, for example "emigrated to the US or did not emigrate to the US", whose occurrence is investigated while controlling for other, independent, dichotomous variables. That is, the "effect" of an independent variable, which is measured by its coefficient, represents the odds ratio of the potential to influence the dependent variable – the event of emigration to a particular destination, while all the other variables are held constant. The advantage of using logistic regression is the ability to examine the effect of any one factor while controlling for the effects of other factors. In addition, the stepwise method was used, which chooses the independent (explanatory) variables that have the most influence on the dependent (explained) variable by a process of progressively entering each potential explanatory variable into the model according to its statistical significance, and removal of a variable which, at any stage in this process, changes from being significant to being non-significant. The procedure ends when no additional variables are left to be put into the model, and there are no non-significant variables that need to be removed from the 27 model. Details on the variables and the process of carrying out the regression appear later in this chapter. The research population comprising the database of Israeli families outside of Israel includes families of men and women who are or were residents of Israel, listed in the Population Register, with children up to age 17, who have at least one child born outside of Israel between the years 1996–2008. All families were included for which there was an indication in the Population Register records that they had a child or children born outside of Israel, no matter what the marital status of the parents was, or whether both spouses were Israeli residents or whether one was not an Israeli resident. The status "citizen residing abroad" (CRA) is given to children who are born outside of Israel to an Israeli parent who lives predominantly outside of Israel, and whose parent requested that the child be recorded in the Population Register at an Israeli consulate abroad. At the time of registration, the child is assigned an identity number and his or her personal information is entered into the Population Register, and the status "CRA" is applied, as well as the locality code 0000, which indicates that the child does not live in Israel. In the current work, these families were combined into a file entitled "Citizens Residing Abroad" (CRA). At the same time, a record is made of children who were born abroad to Israeli parents, entitled "immigrant citizens" (IC). An IC is a person who was born to an Israeli citizen while he or she was staying abroad, and who entered Israel with the intent of settling (Israel Central Bureau of Statistics, 2010, p. 225). These children born abroad are also an indication that their families lived outside of Israel for a substantial period of time, before the child and/or family returned to Israel. In the current research, these families were aggregated into a file entitled "Immigrant Citizens" (IC). The two files CRA and IC constituted the basis for the new database of Israeli families abroad (for more information on the process of constructing these files, see Appendix A). The following is a detailed description of the construction of the file of the research population. 28 Diagram 1. Procedure for Building the Database of Israeli Families Abroad: CRA IC File 1: Source: Children born abroad during 1996–2008 to an Israeli citizen, and whose status is "CRA". Source: Population Register 04/2009. Number of records: 27,480 File 2: Source: Children born abroad during 1996–2008 to an Israeli citizen, and who immigrated to Israel. Source: Multi-year data file of ICs, based on the Population Register, extracted from the CBS's migration system Number of records: 25,565 File 1b: Constructing administrative families: Adding parents and children through age 17 to the records in File 1. Source: Population Register 04/2009. Number of records: 70,035 File 2b: Constructing administrative families: Adding parents and children through age 17 to the records in File 2. Source: Population Register 04/2009. Number of records: 67,547 Research group File 3: Merging of Files 1b and 2b. Attaching all characteristics from the Population Register 04/2009 to each record. Number of records: 133,539. This file included the entire research group. The variable "country of birth" of the records in File 1 + 2 was copied to the other family members and labelled "destination country". In the first stage, a database of the Israeli families abroad was created. This database contained 133,502 individual records, which constituted 34,047 families that were created during the process of constructing administrative families, in a process that was almost identical to that performed in the 2008 Population Census (see Appendix A). This file included all children born abroad during the research period who were given a status of "CRA" or "IC", the child's parents, and siblings through age 17. This age was chosen as an 29 age before independence according to the law; at this age, the vast majority of children are still under their parents' legal care and migrate together with them. In the second stage, characteristics were added to the records in the file by linking between records in this file to other files with identification information existing at the CBS. These files were: Population Register 4/2009: The determinant date for inclusion of individuals and their characteristics was Dec. 31, 2008. Demographic characteristics from this file were added: sex, age, marital status, religion, country of birth, nationality, status in the Population Register, year of gaining immigration status, year of entry, year of birth, and year marital status took effect. 20% Sample from the 1995 Population Census: From this sample file, which contained 20% of the census population, the following characteristics were added: education, occupation, and marital status at the time of the census. This was done for 6,400 records which linked between File 3, mentioned above, and the sample file of the census. CBS database of degree recipients 1983–2008: This database contains data on academic degrees and other post-secondary certificates, the field in which the degree was received, the year of receipt of the degree, and the institution granting the degree. The following characteristics were added from this database: type of degree, year of degree, and field of the degree. There were 15,188 records from File 3 that were linked to this database. Demographic Management of the Population by Individual Records (DMPIR) during 1996– 2008: This system documented the changes in annual population components (entrances/exits from the population and other changes in status) at the level of the individual. From this system, characteristics of entrance and exits from the population, as well as year of exit to abroad, were added. The following numbers of records were linked, with these statuses: past immigrants to Israel (27,330 records linked), family unification (2,457 records linked), Israelis who stayed abroad and returned to Israel, termed "returning 31 residents" (12,952 records linked), Israelis staying abroad who did not yet return (23,929 records linked). Multi-year file of immigrants to Israel 1984–2008: A file including all persons who immigrated to Israel and received immigrant status during these years; a file of individuals records. From this file, the characteristic of last country of residence before immigration was added to 13,947 records that linked between File 3 and this file. Appendix A contains more information on the process of linking the files and its results. After adding the characteristics from all of these files with identification information, a database was created of the characteristics of the families investigated that included 55 variables from various files that were linked to the base file that was created based on the Population Register. Some of the variables were taken as is and some underwent processing to become calculated variables that were prepared especially for the purpose of characterizing the records in the database. Description of the Variables Included In the New Database on Emigrating Families Status variables: the source of the record, the source of the record in the process of constructing the administrative family, status in the Population Register, and various variables denoting the different linkage statuses of the specific record with regard to the files added to the database. Demographic variables: age, sex, religion, nationality, country of birth, year of immigration, last country of residence before immigration, marital status, year marital status took effect, number of persons in family, number of children in family, and relationship to the household representative. Socio-economic variables: occupation in the 1995 Census, participation in the annual civilian labour force, highest degree in the 1995 Census, degree in the CBS's database of academic degrees, and field of study in the CBS's database of academic degrees. 31 In the third stage, checks were made of the quality of the database of Israeli families abroad that was created in the process described above. First, processing was carried out to locate the records from the family database in the CBS's statistical databases called "possible emigrants". This refers to three databases: (a) a stock found in the DMPIR, which includes a stock of persons who might possibly have emigrated during the years 1995–2008, for each year; (b) a "surplus" database from the 1995 Population Census, consisting of records that were in the Population Register in 1995 but were not found in the field at the time of the census enumeration, and (c) the database "possible emigrants, from the 2008 Census", which is based on data from various sources, such as border control, National Insurance, and individuals who were not enumerated in two previous censuses. In these three databases of possible emigrants, together, 81,189 records were found from File 3. These records constitute 61% of all records in the database that was created. This percentage can be explained by the fact that some of the records in the series are those of families of "ICs" who stayed abroad for a period of time and then returned to Israel, and therefore were removed from the databases of "possible emigrants". This did not prevent such records from appearing in the database because they stayed abroad for a substantive amount of time in the past, and their choice to emigrate to a particular destination is relevant to the research goals. For details of the comparison between the database to the database of "possible emigrants", see Appendix A. Second, the question arose as to whether the research population included in the current database of emigrating families appropriately represents, with sufficient reliability, the entire population of emigrating Israeli families. To address this question, two different examinations were made during the research process of the reliability of the methodology. The examinations focused on comparing two groups with similar characteristics that were derived from two different sources. First, a comparison was made between the database and the United States 2000 Census. Similar populations were derived from these two sources consisting of families of only parents and children, with children up to age 18, and at least 32 one parent a native Israeli and one child born in the US. Comparison of these two files showed great similarity in a number of the characteristics examined such as average family size, age structure, and holders of first and second degrees. Second, a comparison was made of the field of academic study of native Israelis who were holders of third degrees (PhDs) whose destination country was the US, between the new database and the results of a survey that was published as part of the report of the Shohat Commission on Higher Education in Israel (Council for Higher Education, 2007). Again, the results showed great similarity between the characteristics of the two groups. These examinations increase the likelihood that the quality of the data in the series is good and reliably represents the population of families emigrating from Israel. For details on the reliability checks of the database, see Appendix B. Third, the use of the CBS's database of degree recipients involves a certain upward bias in the data of education level because most of those in it are recipients of academic degrees, and only a minority are recipients of non-academic degrees. Although these data were added to the data on highest certification in the 1995 Census, in which the sample has equal representation to non-academic certificate holders, there is still apparently underrepresentation in the research group of persons with education at a level lower than academic. The problem of bias is lessened due to several reasons: As stated, the 1995 Census data reduce the bias in data from the database of degree recipients; examination of the education data among the research group from the Census only, by the three selected emigration destinations, revealed that the gaps in educational level among the destinations were maintained in this data also; the distribution of the academic degrees over the emigration destinations has much variance in and of itself, and therefore is suitable for testing the research hypotheses that focus on examination of emigration of persons with relatively high skill levels. Fourth, the data available for this study did not include education acquired by the immigrants to Israel before they came to Israel. The CBS does not have data on education of immigrants in the format of "highest certificate", the variable that was 33 used for persons in this study who acquired their education in Israel. Therefore, in the general database, it is possible that there is under-representation of the skills of persons with higher education who immigrated to Israel and then proceeded to other destinations or returned to their country of birth. Nonetheless, in the research group, upon which the hypotheses were tested, only families for whom there was education data for at least one of the spouses were included. Testing the Research Hypotheses To allow testing of the research hypotheses, a smaller research group that met the required criteria for testing the research hypotheses was derived from the broader database of families of Israelis abroad. This research group included families who met the following criteria: Married parents, both spouses appearing in the Population Register as spouses of one another, and at least one spouse for whom there was education data (from the 1995 Census or from the CBS database of degree recipients 1983–2008) or occupational data (from the 1995 Census). The purpose was to base the testing of hypotheses on a population with known data. Children were not included in this file because it was assumed that, as minors going along with the parents, they did not influence the choice of destination. Educational and/or occupational data constitute a representation of the level of skills of the couple, which, according to the research hypothesis, greatly influences the choice of emigration destination. These variables were used based on the wide consensus in the literature that deals with migration that education acts as a proxy for an individual's level of skills (see, for example, Borjas, 1994; Cohen, 2009; Ouaked, 2002; Skeldon, 2009; United Nations, 2009b). The separate records for the spouses were merged into one record, with each couple having some variables shared by both spouses and some individual variables for each spouse. At the end of this stage, the regression file included 10,695 records (couples), which included data on 21,390 individuals. 34 Sources of Data In testing the research hypotheses, data from several sources was used, and variables were calculated based on these data. The data sources were: The Population Register: Derived for Dec. 31, 2008 (received by the CBS in April 2009). The main source for identification and for creation of the base of the database of Israeli families abroad. The data on "country of birth" in the Population Register is considered very reliable. The families were created using an algorithm of construction of administrative families that was developed for the 2008 Population and Housing Census, with one difference: there was no importance to the geographic location of the family in Israel, and therefore there was no meaning to the concept of "household", rather, only to the creation of an "administrative family" with familial relations of parents to children. This avoided the need to use data on address of residence, which has relatively low reliability in the Register. As part of creating the file of married couples and their skills, only in cases in which only one spouse had data on "identification number of spouse" was the missing number copied to the other spouse (there were 44 such records). In addition, when the listed identification number of a spouse was not found in the Register, apparently because of errors in recording, the "identification number of spouse" was deleted and the status was changed to marriage without a spouse in the Register (there were 217 such cases). 20% Sample from the 1995 Population Census: The data "highest degree" and "occupation" were taken. From the records sampled in the Census, 4,910 records were also found in the file of married couples and their skills that was built in this project. CBS's Database of Degree Recipients: Created and maintained by the Education Department of the CBS. The database includes persons who received their highest degree from academic institutions in Israel, based on the following criteria: From the universities and the Open University during 1989–2003, from the academic colleges during 1993–2003, and from colleges of education during 1995–2003. Data were missing on recipients of degrees from the academic colleges in 2001 and from the Open University in 2003. From this 35 database, 13,650 degree recipients were also found in the file of married couples and their skills that was built in this project. The research hypotheses were tested as follows. To test the hypotheses, that is, to separately examine groups of families that emigrated to each of the three defined destinations, statistical analyses were performed using three different methods: (a) distributions of variables using descriptive statistics; (b) stepwise binary logistic regression on all the existing variables, with the goal of identifying only the significant variables that optimally contributed to characterizing the families emigrating to a particular destination, and (c) binary logistic regression on major selected variables, to also enable analysis and explanation of non-significant variables and their directions of influence. Stages of Implementing the Research Method: 1) Each couple in the research file was classified into one of four groups of destination countries, based on the model built in the research hypotheses: the United States (as mentioned, Canada was not included together with the US in this model, even though it is an emigration destination for many Israelis, because the level of economic and employment returns that characterize the US are higher than that in the Canada economy); Central and Western European countries, including: the United Kingdom, Germany, France, Belgium, Netherlands, Switzerland, Italy, Spain, Denmark, Austria, Ireland, Sweden, Norway, Finland, and Luxembourg; the former USSR and Eastern European countries, including: the Russian Federation, Ukraine, Hungary, the Czech Republic, Romania, Poland, Belarus, Kyrgyzstan, Bulgaria, Latvia, Kazakhstan, Georgia, Azerbaijan, Serbia, Slovakia, Uzbekistan, Lithuania, Estonia, and Croatia; and a group called Other. The "Other" group was used only in the descriptive statistics, in order to allow characterization of the entire investigated group. Separate models were built for each destination based on the research hypotheses – according to which emigrants with relatively high levels of skills will tend to emigrate to the US; the emigrants to Central and Western Europe 36 will tend to be those with less high skill levels and/or those who are natives of the countries they return to; and emigrants to a republic of the former USSR or to Eastern Europe will tend to have relatively low skill levels and/or be natives of the countries they return to – based on the differing characters of the economies and labour markets in each of the chosen emigration destinations. 2) Variables were chosen that were likely to influence the decision to emigrate and the choice of destination. These variables were: religion, origin, migration history, education, and occupation. (See details below.) 3) The characteristics of the families which emigrated to each of the groups in the model were investigated using three methods of statistical analysis: (a) descriptive statistical analyses of selected variables that demonstrate the differences in the characteristics of the family groups; (b) binary logistic regression using the stepwise method; and (c) binary logistic regression on selected variables, even if not significant, to examine their strengthening or weakening contribution to the model for emigration to a specific destination. Each group of families that emigrated to a particular destination was contrasted in comparison to emigration to all of the other destinations combined. Results of the statistical analyses are presented in the Results section. The three models of emigration destinations that were examined in the statistical regressions demonstrate the logit function of the probability of an event that has a binary outcome, as a linear function of the explanatory variables. The model was as follows: p(emigration i ) 0 1 * X 1 ... n * Xn ln 1 p ( emigration i ) emigrationi - indicates the emigration to a particular destination 37 Each emigration destination (emigrationi) was assigned the value 1 when emigration occurred to it and 0 when emigration did not occur to it. In this manner, the probability of emigration to each of the emigration destinations was calculated in accordance with the explanatory variables. Model A H0 – immigration did not occur to the US H1 – immigration did occur to the US Model B H0 – immigration did not occur to the countries of Western and Central Europe H1 – immigration did occur to the countries of Western and Central Europe Model C H0 – immigration did not occur to republics of the former USSR or to the countries of Eastern Europe H1 – immigration did occur to republics of the former USSR or to the countries of Eastern Europe The variables that were included in the testing of the research hypotheses were based on the literature. Selection of the independent variables to be included in testing the research hypotheses was based on the theories as well as on empirical research. The theories of neoclassical economics and self-selection discuss migration patterns that are influenced by the returns expected in exchange for the skill level of the migrant, who is the rational actor in the economic sphere. In these theories as well as in empirical research, education and occupation act as a proxy for level of measurable skills or level of human capital. The literature review discussed the following research that treated education and occupation as representative of 38 migrants' skill levels: Borjas, 1994; Cohen, 1989, 2009; Eisenbach, 1989; Gould and Moav, 2007; Michaeli, Epstein, and Kheimets, 2007; Skeldon, 2009; and Toren, 1980. Many empirical research studies take into account the migration history of an emigrant or his or her country of birth. This is because in every wave of migration, a not insignificant portion of the migrants are returning to their country of origin, as happens, for example, to Israelis who are natives of Europe or America who return to their countries of birth, or Israelis born abroad who stay only a short time in Israel and then proceed to another destination country, as happened to a portion of the immigrants from the former USSR. The phenomenon of migration of emigrants who were formerly immigrants, who use the host country as a stepping stone to reach another country, is referred to as "transit migration" or "onward migration". In order to investigate the family's country of origin, which can indicate ease of migration back to that country, due to one spouse or a spouse's parent(s) having born in that country, one can use parents' country of birth in order to attempt to explain emigration to a particular country. The literature review discussed the following research that dealt with the migration history of the migrants: Cohen, 2008, 2009; Gould and Moav, 2007; Michaeli, Epstein, & Kheimets, 2007; and United Nations, 2009a. Further, regarding investigation of the variable religion, research that investigated migration of Israelis as well as even CBS data indicate that the emigration rate among nonJewish immigrants from the former USSR is more than double that among Jewish immigrants (Michaeli, Epstein, & Kheimets, 2007). In addition, the research mentions that it is possible that emigration from Israel can be characterized as bi-polar: highly educated persons on one hand, and less educated Israelis (including Arabs) on the other hand (Cohen, 1989). However, even though the religion of the emigrants is considered important when discussing Israel, more comprehensive research does not exist on characteristics of emigration of Israelis by religion. The current research adds information about this topic. 39 A description of the variables follows. Dependent variable: Grouped emigration destination countries. Independent variables: Religion: A variable indicating "both spouses are Jewish" or the opposite value indicating "at least one spouse is not Jewish". Variable at the couple level. Country of birth: A variable indicating "both members of the couple were born in Israel" or the opposite value indicating "at least one member of the couple was not born in Israel". Variable at the couple level. Migration history: "At least one spouse was born in the family's emigration destination", or "the man previously had immigrant status", or "the woman previously had immigrant status". Variables at the couple level and the individual level. Educational and occupational skills: Education – Highest degree: Man has first degree; woman has first degree; man has second degree; woman has second degree (second degree also includes the medical degree M.D.); man has third degree; woman has third degree; man has other education (less than academic); woman has other education (less than academic); man studied a "high scientific" field of study (i.e., a field of academic study that requires a person to have a high level of skills to be accepted and complete the degree; the field can be easily adapted to the international labour market; includes medicine, mathematics, statistics, computer science, physical sciences, biological sciences, engineering, and architecture) or woman studied a "high scientific" field of study. 41 Occupation – In the 1995 Census: Man has an academic professionals or associate professionals and technicians (requiring a high level of skills) occupation; woman has an academic professionals or associate professionals and technicians occupation. Full details on all the variables and all their values that were prepared for the investigation of the research group are presented in Appendix C. The results of the regression analyses are presented in the Results section. 41 3. Results The statistical analyses showed that some of the research hypotheses were confirmed and some were not. Briefly, it was found that, in accordance with the research hypothesis, families that emigrated to the US did have very high levels of skills. In contrast, support was not found for the research hypothesis that the families who emigrated to Central and Western Europe would also be characterized by high levels of skills. Instead, it was found that skills taking the form of a man with a "high scientific" field of study, a woman with a third degree, or a woman having an academic or professional or technical occupation lowered the odds ratio of emigration to this destination. The results additionally showed that the characteristics that better characterized this group were whether one of the couple was born in the emigration destination (return migration), if at least one of the couple was not born in Israel, or if at least one of the couple was not Jewish. In accordance with the research hypothesis, a relatively low level of skills was found among families who emigrated to the former USSR and Eastern Europe. Against the research hypothesis that most of these emigrants would be persons who were born in these countries, it was found that natives of these countries did not tend to return to their countries of origin, but rather to continue to emigrate to more remunerative destinations. The results of the statistical analyses follow. 42 A) Descriptive Statistics Selected Statistical Distributions Diagram 2. Religious Composition of Married Couples, by the Selected Emigration Destinations, in Percentages Both spouses Jewish At least one spouse not Jewish 100.0 90.0 80.0 70.0 60.0 50.0 40.0 30.0 20.0 10.0 0.0 US Central and Western Europe USSR (former) and Eastern Europe Other* * The "Other" group included 1,782 couples who emigrated to all other emigration destination countries, which were not included in the three destinations that were studied. This group is too heterogeneous in terms of the characteristics of the countries included in it to be characterized as a single entity. It is presented for the sake of completeness and in order to present the entire research group. Analysis of the results showed that both spouses were Jewish in approximately 95% of the families that emigrated to the US, and only approximately 5% of these families had at least one non-Jewish spouse. In comparison, the percentage of families with at least one nonJewish spouse increased to approximately 14% among families who emigrated to Central and Western Europe, and was even larger (23%) among the families who emigrated to the former USSR and Eastern Europe. This finding goes along with the increase in the nonJewish population that emigrated to Israel in the large wave of immigration during the 1990s, and the increase in the right to immigrate due to the Law of Return, as well as perhaps the presence of a weaker connection to Israel among this population. 43 Diagram 3. Country of Birth of the Couple, by the Selected Emigration Destinations, in Percentages At least one spouse born in emigration destination country Neither spouse born in emigration destination country 100.0 90.0 80.0 70.0 60.0 50.0 40.0 30.0 20.0 10.0 0.0 US Central and USSR (former) Western Europe and Eastern Europe Other The percentage of emigrating families who emigrated to a destination that was new for both spouses was greater than the percentage of families who migrated back to the country of birth of at least one spouse. This was true for all the destinations. Among the families who migrated back to the country of birth of at least one spouse, the largest percentage (approximately 45%) migrated back to the republics of the former USSR or Eastern European countries. Approximately 29% of the couples who emigrated to Central and Western Europe were returning to the country of birth of at least one spouse, and only 15% of the families who emigrated to the US had come from there. 44 Diagram 4. Origin (Country of Birth of the Couple's Parents), by the Selected Emigration Destinations, in Percentages Couple's parents not born in emigration destination country At least one of couple's parents born in emigration destination country 100.0 90.0 80.0 70.0 60.0 50.0 40.0 30.0 20.0 10.0 0.0 US Central and Western Europe USSR (former) and Eastern Europe Continuing the trend presented in Diagram 3, Diagram 4 shows that over half (52%) of the couples who emigrated to the former USSR or Eastern Europe were from this origin. That is, at least one of the four parents of the couple was born in former USSR or Eastern Europe. Approximately one-fifth (19%) of the couples who emigrated to Central or Western Europe were born to parents who were born in these countries. This European origin allows persons who were born in a European Union country to work in any of these countries, which could motivate a decision to emigrate to these destinations. The lowest percentage of couples whose country of origin was identical to their emigration destination was found for the US: Only 7% of couples who emigrated to the US were children of parents who were born in the US. 45 Diagram 5. Israeli-Born Couples, by the Selected Emigration Destinations, in Percentages USSR (former) and Eastern Europe 3% Other 15% Central and Western Europe 18% US 64% Among families in which both spouses were born in Israel, there was a clear preference for emigration to the US. Only slightly over one-quarter of these families emigrated to Central and Western Europe, and a very small minority emigrated to the former USSR and Eastern Europe. Apparently, when there is no family to return to in the destination country as a social network, many families prefer to emigrate to a destination with relatively high opportunities for returns. 46 Diagram 6. New Immigrants (Immigrated to Israel During 1996–2008) With a High Level of Education (Second Degree or Higher) Who Were Born in the Former USSR or Eastern Europe, by the Selected Emigration Destinations, in Percentages Other 20% USSR (former) and Eastern Europe 15% US 48% Central and Western Europe 17% Almost half of the new immigrants born in the former USSR or in Eastern Europe with a relatively high level of education (second degree or higher) who emigrated chose to emigrate to the US rather than to their country of origin, even though they likely had social networks in their country of origin. Including those who emigrated to Central and Western Europe, the percentage of immigrants born in the former USSR or in Eastern Europe who chose to emigrate to a more remunerative economy reached 65%. Only 15% migrated back to their country of origin. These findings strengthen the research hypothesis that predicted that the higher the skill level, the greater the tendency to emigrate to a destination where the skills are better remunerated. 47 Diagram 7. Education of Men and Women by Highest Degree, by the Selected Emigration Destinations, in Percentages US Central and Western Europe USSR (former) and Eastern Europe Other 100 90 80 70 60 50 40 30 20 10 0 Less than academic B.A. M.A. M.D. PhD At all levels of education, more than half the families emigrated to the US. This was most pronounced among holders of PhD degrees, of whom over 80% emigrated to the US, and among holders of a medical degree (M.D.), of whom over 70% emigrated to the US. The smallest percentage who emigrated to the US was found among those with less than an academic education, relative to the other levels of education. Central and Western Europe was in second place in terms of attractiveness to those with a higher level of education, followed by destinations in the former USSR and Eastern Europe. 48 Diagram 8. Education of Men and Women With a Second Degree or Higher, by the Selected Emigration Destinations, in Percentages USSR (former) and Eastern Europe 3% Other 14% Central and Western Europe 16% US 67% This diagram also shows that emigrants with the highest levels of education emigrated to the US. In comparison, only one-quarter as many emigrants went to Central and Western Europe, and even fewer emigrated to the former USSR and Eastern Europe. Diagram 9. Men and Women With "High Scientific" Fields of Study, by the Selected Emigration Destinations, in Percentages USSR (former) and Eastern Europe 2% Other 13% Central and Western Europe 14% US 71% The emigration of those with high skill levels is also seen in Diagram 9. A degree in one of these fields of study (in the current study: medicine, mathematics, statistics, computer 49 science, physical sciences, biological sciences, engineering, or architecture) can be more easily adapted to the needs of the US labour market. The trend was the same as that indicated by the previous diagram, which showed those with second and higher degrees: about 70% of them emigrated to the US. Diagram 10. Men and Women Who Were Academic Professionals or Associate Professionals and Technicians in the 1995 Census, by the Selected Emigration Destinations, in Percentages Out of Total Emigrants in the Research Group With These Occupations Approximately 62% of the emigrants who were Academic professionals or Associate professionals and technicians in 1995 emigrated to the US, 16% to Central and Western Europe, and 6% to republics of the former USSR or to Eastern Europe. These occupations are usually characterized by a high level of skills, together with easier adaptation to the needs of the US labour market. These factors explain the large share of the US in attraction of emigrants with these skills. 51 Summary Table: Descriptive Characteristics of the Research Group, by Emigration Destinations Variable Name Emigration Destination United Central and USSR Other States Western )former) and Europe Eastern Europe Percentages Religious composition of married couples: Both spouses Jewish At least one spouse not Jewish 3..9 3..9 6... 3..9 7.6 79.6 99.7 3.3 3..9 67.9 ...9 3..7 77.3 93.6 77.3 77.. 39.9 37.9 73.. ..3 73.7 .9.. .7.. 73.. 9.. Country of birth of couple: Neither spouse born in emigration destination country At least one spouse born in emigration destination country Origin (country of birth of the couple's parents): Couple's parents not born in emigration destination country At least one of couple's parents born in emigration destination country Both spouses born in Israel 7... 51 Variable Name Emigration Destination United Central and USSR Other States Western )former) and Europe Eastern Europe 73.. 76.. 7... 9... Less than academic .7.. 99.9 ..3 76.3 B.A. ...9 73.6 9.9 7..3 M.A. .7.3 73.6 9.. 7..3 M.D. 69.7 6.. ..3 73.9 PhD 63.7 79.. 7.. 6.. Men and women with second degree or .6.. 7... 9.. 77.. 67.. 77.. 9.. 79.. academic professionals .7.3 77.7 ... 7..9 associate professionals and .3.6 73.7 ..7 7..9 New immigrants (immigrated during 1996–2008) with second degree or higher, born in former USSR or Eastern Europe Education of men and women, by highest degree: higher Men and women with "high scientific" fields of study Men and women with academic professionals or associate professionals and technicians occupation in the 1995 Census: technicians 52 B) Statistical Analyses: Regression Results Below is a detailed description of the results of the regression analyses of emigration to each of the three destinations. Two procedures involving binary logistic regression were used. The first was a binary logistic regression analysis which included variables chosen in the stepwise procedure out of all the relevant variables in the file. The second was a binary logistic regression analysis run on selected variables (i.e., there were two regressions for each destination). The second regression analysis was performed to investigate the extent and direction of the contribution to the model of relevant variables, even if not significant, thus helping to strengthen or disprove the research hypotheses. It should be emphasized that in accordance with the research hypothesis, which regarded the characteristics of the emigrants to each emigration destination separately, each model stands on its own and the regression was built comparing the families who emigrated to a particular destination to all other families in the file. 53 Table 1.1: Results of Stepwise Regression Analysis Analyzing the Chances of Emigrating to the US (Model A) Independent Variables Estimate Odds Ratio Estimate Both spouses Jewish ..376*** 9..63 Both spouses born in Israel ..7.7*** 7..3. At least 1 spouse born in family's emigration ..99.** 7.9.. Man has "high scientific" academic study field ...73*** 7.377 Woman has "high scientific" academic study field 0.213* 7.993 Man has PhD ..76.*** 7...3 Woman has PhD ..69.*** 9.77. Man has B.A. -0.164** ..373 Woman has B.A. ..97.*** 7.977 c (area under the ROC curve) ...7. Number of observations (couples) H1 .69.3 destination country * p < .01; ** p < .001; *** p < .0001. The results of the analysis in Table 1.1 (Model A) show that the statistically significant characteristics for the population of families that emigrated to the US were families in which both spouses are Jewish, and in which both spouses were born in Israel or alternatively at least one of the spouses was born in the US. In addition, a high level of skills, measured as a "high scientific" academic study field or as a PhD degree, for the man or the woman, strengthens the tendency to emigrate to this destination. The man's holding a B.A. degree has a negative, significant influence on the tendency to emigrate to the US, whereas the woman's holding a B.A. degree has a positive, highly significant influence. 54 Table 1.2: Results of Regression Analysis Analyzing the Chances of Emigrating to the US (Model A, on Selected Variables) Odds Ratio Independent Variables Estimate Estimate Both spouses Jews ..377*** 9..67 Both spouses born in Israel ..7.9*** 7..67 ..993** 7.96. Man had immigrant status in the past -..7.3 ..377 Woman had immigrant status in the past -...7. ..33. ...99*** 7.339 ..97.* 7.977 Man has PhD ..739*** 7..9. Woman has PhD ..673*** 9.77. Man has M.A. or M.D. ....6 7..6. Woman has M.A. or M.D. ..7.9 7.7.3 Man has B.A. -0.153* ..3.3 Woman has B.A. ..9..*** 7.993 Man has less than an academic education -...63 ..397 Woman has less than an academic education ...73 7..9. Man has an academic professionals or associate professionals and technicians occupation ..793 7.7.. Woman has academic professionals or associate professionals and technicians occupation -....9 ..393 c (area under the ROC curve) ...76 Number of observations (couples) H1 .69.3 At least 1 spouse born in family's emigration destination country Man has "high scientific" academic study field Woman has "high scientific" academic study field * p < .01; ** p < .001; *** p < .0001. 55 Table 1.2 displays the results of the logistic regression analysis for families that emigrated to the US, with the explanatory variables including variables relating to religion, origin, migration, and skills. The results of the regression show that the variable that contributes the most to predicting emigration to the US is both spouses being Jewish (this raised the odds ratio by 2.6 times; p < .0001). The next variables, in decreasing order of contribution to the prediction (according to the Wald chi-square), were: man has "high scientific" academic study field (this raised the odds ratio by 1.9 times; p < .0001), both spouses born in Israel (this raised the odds ratio by 1.6 times; p < .0001), woman has a PhD (this raised the odds ratio by 2.1 times; p < .0001), man has a PhD (this raised the odds ratio by 1.6 times; p < .0001); woman has a B.A. (this raised the odds ratio by 1.2 times; p < .0001), at least one spouse was born in the US (this raised the odds ratio of return migration to the US by 1.3 times; p < .001), woman has a "high scientific" academic study field (this raised the odds ratio by 1.2 times; p < .01), and man has a B.A. (this lowered the odds ratio by 0.9 times; p < .01). Additional variables that were included in the model but were not found significant and therefore their contribution to the model is limited, were as follows, in decreasing order by the extent of their contribution to the model and direction of influence: man had immigrant status in the past (lowered the odds ratio), woman has an M.A. (raised), man has M.A. (raised), man has less than academic education (lowered), woman has less than academic education (raised), man has an academic professionals or associate professionals and technicians occupation (raised), woman has an academic professionals or associate professionals and technicians occupation (lowered), and woman had immigrant status in the past (lowered). These findings indicate that the following variables significantly strengthen the odds ratio for emigration to the US: if both spouses are Jewish and born in Israel, if at least one of them was born in the US, if they have high-level skills measured as a PhD (held by the man or the woman) or by a "high scientific" field of study (more true for the man than the woman). There 56 is a dual trend for holders of B.A.s: When the woman has a B.A., the tendency to emigrate to the US is stronger, and when the man has a B.A., the tendency is weaker. Table 2.1: Results of Stepwise Regression Analysis Analyzing the Chances of Emigrating to Countries of Central or Western Europe (Model B) Odds Ratio Independent Variables Estimate Estimate At least 1 spouse not Jewish At least 1 spouse not born in Israel At least 1 spouse born in family's emigration destination country Man has "high scientific" academic study field Woman has PhD ...9.*** 7.39. ..739* 7.77. ..939*** 7.963 -0.597*** ....3 -0.467* ....9 c (area under the ROC curve) ...7. Number of observations (couples) H1 969.7 * p < .01; *** p < .0001. Analysis of the results shown in Table 2.1 (Model B) shows that families with the following characteristics have a statistically significant higher tendency to emigrate to Central and Western Europe: at least one spouse not Jewish, one spouse was not born in Israel or at least one spouse was born in the emigration destination country. Regarding skills, a high level of skills that takes the form of either the man having a "high scientific" academic study field or the woman having a PhD, negatively influences emigration to this destination. That is, the analysis shows that holders of these types of skills did not tend to choose emigration to this destination. 57 Table 2.2: Results of Regression Analysis Analyzing the Chances of Emigrating to Countries of Central or Western Europe (Model B, on Selected Variables) Estimate Odds Ratio Independent Variables Estimate At least one spouse not Jewish ...33*** 7.39. At least one spouse not born in Israel ..976** 7.77. At least 1 spouse born in family's emigration destination country ..97.** 7.963 Man had immigrant status in the past -0.108 ..336 Woman had immigrant status in the past -0.134 ..367 -0.581*** ....3 Woman has "high scientific" academic study field -...39 ..37. Man has PhD -....3 ..399 Woman has PhD -..779 ....9 Man has M.A. or M.D. -...33 ..377 Woman has M.A. or M.D. -...6. ..396 Man has B.A. ....7 7...9 Woman has B.A. -..... ..3.7 Man has less than an academic education ...9. 7..9. Woman has less than an academic education -...97 ..3.3 Man has an academic professionals or associate professionals and technicians occupation -..79. ..363 Woman has an academic professionals or associate professionals and technicians occupation -..993 ..63. c (area under the ROC curve) ...79 Number of observations (couples) H1 969.7 Man has "high scientific" academic study field ** p < .001; *** p < .0001. Table 2.2 displays the results of the logistic regression analysis for families which emigrated to countries of Central and Western Europe, with the explanatory variables including 58 variables relating to religion, origin, migration, and skills. The results of the regression show that the variable that contributes the most to predicting emigration to this destination is the man having a "high scientific" academic study field (this lowered the odds ratio by 0.6 times; p < .0001). The next variables, in decreasing order of contribution to the prediction (according to the Wald chi-square), were: at least one spouse not being Jewish (an opposite and complementary variable to the variable "both spouses Jewish"; raised the odds ratio by 1.8 times, p < .0001), at least one spouse not born in Israel (an opposite and complementary value of the variable "both spouses born in Israel"; raised the odds ratio by 1.4 times, p < .001), and at least one spouse born in the emigration destination country (raised the odds ratio by 1.3 times, p < .001). Additional variables that were included in the model but were not found significant and therefore their contribution to the model is limited, were as follows, in decreasing order by the extent of their contribution to the model and direction of influence: woman has a PhD (lowered the odds ratio), woman had immigrant status in the past (lowered), woman has an academic professionals or associate professionals and technicians occupation (lowered), man had immigrant status in the past (lowered), man has M.A. (lowered), woman has a "high scientific" academic study field (lowered), man has B.A. (raised), man has an academic professionals or associate professionals and technicians occupation (lowered), woman has an M.A. (lowered), woman has a B.A. (lowered), man has a PhD (lowered), man has less than academic education (raised), and woman has less than academic education (lowered). These findings indicate that, in accordance with the research hypothesis, if at least one spouse was born in the emigration destination country, the odds ratio for emigrating to that destination increase significantly. In addition, the variables "at least one spouse not Jewish" and "at least one spouse not born in Israel" were found to be highly significant and reliably explained emigration to this destination. Against the research hypothesis, a high level of skills was not a good predictor of emigration to this destination. Most of the education variables were not significant, and even weakened (had a negative influence on) the 59 tendency to emigrate, according to the odds ratios. Note that the man having a PhD significantly weakened the tendency to emigrate to this destination. Religion and origin variables were found to be better predictors of emigration to this destination. Table 3.1: Results of Stepwise Regression Analysis Analyzing the Chances of Emigrating to Republics of the former USSR or to Eastern European Countries (Model C) Estimate Odds Ratio Independent Variables Estimate At least one spouse not Jewish 7.777*** 9..93 At least one spouse born in family's emigration -0.338* ..679 destination country status in the past Man had immigrant ...79*** 7.37. Woman had immigrant status in the past ..936** 7.733 Man has "high scientific" academic study field -0.626*** ...97 Woman has "high scientific" academic study field -0.794** ..7.9 Woman has B.A. -0.550*** ...6. Man has less than an academic education ..97.* 7.777 Woman has an academic professionals or associate professionals and technicians occupation ...39** 9...7 c (area under the ROC curve) ...33 Number of observations (couples) H1 777 * p < .01; ** p < .001; *** p < .0001. An analysis of the results of Model C shows that the characteristics that increase the tendency to emigrate to the former USSR and Eastern Europe are as follows: At least one spouse is not Jewish or the man or woman had immigrant status in the past. Conversely, if at least one spouse was born in the emigration destination country, the odds ratio for emigration to this destination decreases. Regarding skills, the following have a significant, negative influence on emigration to this destination: the man or woman having a "high scientific" academic study field or the woman having a B.A. The man having less than an academic education or the woman having an academic professionals or associate 61 professionals and technicians occupation significantly increase the tendency to emigrate to this destination. Table 3.2: Results of Regression Analysis Analyzing the Chances of Emigrating to Republics of the former USSR or to Eastern European Countries (Model C, on Selected Variables) Estimate Odds Ratio Independent Variables Estimate 7..39*** 9.33. At least one spouse not born in Israel -0.088 ..37. At least 1 spouse born in family's emigration -0.360 ...33 ...66*** 7.3.3 ..77.* 7..7. -0.646*** ...97 -0.649* ...99 Man has PhD -...33 ....9 Woman has PhD -..6.. ..737 Man has M.A. or M.D. -..9.3 ..377 Woman has M.A. or M.D. -..9.3 ..667 Man has B.A. -....7 ..333 -0.703*** ..73. Man has less than an academic education ..933 7.99. Woman has less than an academic education ..979 ..697 ....7 ..333 ..3.6*** 9.937 At least one spouse not Jewish destination country Man had immigrant status in the past Woman had immigrant status in the past Man has "high scientific" academic study field Woman has "high scientific" academic study field Woman has B.A. Man has an academic professionals or associate professionals and technicians occupation Woman has an academic professionals or associate professionals and technicians occupation c (area under the ROC curve) Number of observations (couples) H1 ..6.9 777 * p < .01; *** p < .0001. 61 Table 3.2 displays results of the logistic regression for families who emigrated to the republics of the former USSR or to Eastern European countries, with the explanatory variables including variables relating to religion, origin, migration, and skills. The results of the regression show that the variable that contributes the most to predicting emigration to this destination is when at least one spouse is not Jewish (this raised the odds ratio by 3.0 times; p < .0001). The next variables, in decreasing order of contribution to the prediction (according to the Wald chi-square), were: whether the woman had a B.A. (lowered by 0.5 times; p < .0001), the man having immigrant status in the past (raised by 2.0 times; p < .0001), whether the man had a "high scientific" academic study field (lowered by 0.5 times; p < .0001), the woman having an academic professionals or associate professionals and technicians occupation (raised by 2.3 times; p < .0001), the woman having immigrant status in the past (raised by 1.5 times; p < .01), and whether the woman had a "high scientific" academic study field (lowered by 0.5 times; p < .01). Additional variables that were included in the model but were not found significant and therefore their contribution to the model is limited, were as follows, in decreasing order by the extent of their contribution to the model and direction of influence: at least one spouse born in the emigration destination country (lowered the odds ratio), woman has less than an academic education (raised), man has less than an academic education (raised), man has a PhD (lowered), woman has a PhD (lowered), man has an M.A. (lowered), at least one spouse was not born in Israel (lowered), man has B.A. (lowered), and man has an academic professionals or associate professionals and technicians occupation (lowered). The results of the regression show that when at least one spouse is not Jewish or the man or woman were immigrants in the past, the odds ratio of emigrating to this destination increases. It is possible that the persons involved are non-Jewish immigrants who immigrated to Israel in the past, but due to various reasons (perhaps also related to their not being Jewish) they did not acclimate well in Israel and therefore tend to migrate back to this 62 destination. In accordance with the research hypotheses, the families who emigrated to the former USSR or to Eastern Europe were characterized by a relatively low level of skills: The man (significant in the stepwise regression) or woman having less than an academic education increased the odds ratio of emigrating to this destination, and the other education variables lowered the odds ratio of emigrating to this destination. It should be noted that whereas the large immigration to Israel from the former USSR in the 1990s was characterized by a relatively high level of education, in the current research, the degrees of persons who earned them outside of Israel were not included in the CBS's database of academic degrees. Therefore, their degrees were included only if they were included in the 20% sample from the 1995 Population Census. Against the research hypothesis, according to which a strong finding was expected to be return migration to the country of birth by persons who were born in a country of the emigration destination, i.e., persons who immigrated to Israel and decided to return to their country of birth, it was found that persons born in these countries did not tend to return to their country of birth but rather to continue to emigrate to more remunerative destinations. 63 4. Discussion and Conclusion The research hypotheses of the current work focused on emigration of Israeli families to three different destinations. The characteristics of the families were examined, along with a research hypothesis about the level of skills characterizing each group of families, taking into account the economic structure characterizing each emigration destination that was examined. In accordance with this, three research hypotheses were formulated: (1) The Israeli families emigrating to the US will tend to be characterized by a very high level of skills; families of US natives who immigrated to Israel will tend to return to their country of origin; and native Israelis will tend to emigrate to the US. (2) The Israeli families emigrating to Central and Western Europe will be characterized by a relatively high level of skills. Families of natives of one of these countries will tend to return to their country of origin, from which they can proceed to any of the European Union countries. (3) The Israeli families emigrating to the former USSR and Eastern Europe will be characterized by a relatively low level of skills. In addition, natives of the former USSR and Eastern Europe will tend to return to their country of origin. Overview of the Findings Emigration to the US The findings that emerged from examination of the skill levels of the spouses, through examining their levels of education and occupations, consistently showed, over all the statistical analyses, that persons with higher educational levels emigrated to the US. In agreement with the research hypotheses, the largest proportion of holders of doctoral degrees (among men and women), holders of medical degrees (men and women), and persons who studied a "high scientific" field of study (men and women), emigrated to the US. These findings were statistically significant. In contrast, it was found that among men, a lower level of education (a first degree [significant] or education lower than academic) weakened the tendency to emigrate to the US. It might be that in the family compositions in 64 the file, there is a tendency for the men to have higher educational levels than their spouses, and in these cases, the emigration destination was chosen based on the man's skills. There was no clear trend regarding the characteristic of occupation: The man's having an academic professionals or associate professionals and technicians occupation increased the tendency to emigrate to the US, but the woman's having such an occupation decreased this tendency (neither of these were significant). Family characteristics that strongly influenced emigration to the US were both spouses being Jewish and both being native Israelis (both significant at a level of p < .0001). Apparently, when both spouses are native Israelis and they do not have a family or social network in an emigration destination (as do persons born outside of Israel), the thought is that once the choice has been made to emigrate, one should emigrate to a destination whose economic structure allows maximal returns. The religion of the couple might be explained by the fact that in Israel, the vast majority of those receiving advanced academic degrees are Jewish (in the academic year 2007/2008, 94.5% of third degree recipients were Jewish, 2.8% were Arab and 2.7% were unclassified by religion or Christians; Israel Central Bureau of Statistics, 2010, p. 436). It is possible that there is an association between these two variables. In accordance with the research hypothesis, it was found that US natives tended to migrate back to the US if they left Israel (significant at a level of p < .001). Central and Western Europe Contrary to the research hypothesis, families who emigrated to Central and Western Europe were not characterized by a high level of skills. The man having a "high scientific" field of study or the woman having a PhD significantly weakened the tendency to emigrate to this destination. In addition, although not significant, the other variables for relatively high educational level (second and third degree, for both sexes, and "high scientific" field of study, for women) negatively influenced the tendency to emigrate to this destination, and variables for lower educational levels (first degree and education less than academic, both for men) had positive influences on the tendency to emigrate to Central and Western 65 Europe. Regarding the occupational variable, the same trend was observed: the man or woman having an academic professionals or associate professionals and technicians occupation weakened the tendency to emigrate to this destination (not significant). The characteristics that best described the emigrants to Central and Western Europe were that at least one spouse was not Jewish, and at least one spouse not having been born in Israel (both significant, at levels of p < .0001 and a level of p < .001, respectively). It might be that the low level of skills found among emigrants to this destination is tied to the characteristic of at least one spouse not being Jewish. As mentioned, in Israel, the educational level of Jews tends to be higher than that of non-Jews. In accordance with the research hypothesis, it was found that among families in which at least one spouse was born in the emigration destination country, the tendency to emigrate to this destination was strengthened (significant at a level of p < .001). Former USSR and Eastern Europe In accordance with the research hypothesis, relatively low levels of skills were found among families who emigrated to the former USSR and Eastern Europe. All of the higher education variables were found to negatively influence the tendency to emigrate to this destination; the man or woman having a "high scientific" field of study and the woman having a first degree were significant. In accordance with the research hypothesis, education less than academic, for both men and women, also positively influenced the tendency to emigrate (although this was not significant). An additional statistical analysis that was performed regarding the education of "new immigrants" who immigrated from the former USSR or Eastern Europe having second or higher degrees showed that almost half of them preferred to emigrate to the US (48%). Fewer preferred to emigrate to Central and Western Europe (17%), and even fewer to their country of origin (15%). Hence, also among immigrants from the former USSR or Eastern Europe, as skill levels increase, so too does the tendency to emigrate to a destination where the skills are remunerated at the highest levels. Emigration to such destinations was preferred by 65% of these emigrants over return to the country of origin. 66 Among the reasons for this, it can be suggested that it might be that improvement in skill level acquired in Israel enables a move to a more attractive destination, or that from the outset, Israel was only a stepping stone on the way to the desired destination. The fact that the man or woman having an academic professionals or associate professionals and technicians occupation increased the tendency to emigrate to this destination (among women, this was even highly significant, p < .0001) might be explained in that these families were part of the large immigration wave, which was characterized by relatively high occupational levels. In accordance with the research hypothesis, it was found that at least one spouse not being Jewish increased the tendency to emigrate to the former USSR or Eastern Europe. Contrary to the research hypothesis, it was found that the tendency to emigrate to this destination was weakened when at least one spouse was born in a country of the family's emigration destination. The reason for this is apparently similar to the explanation mentioned above, according to which the immigrants from these destinations who had relatively high skill levels strongly preferred to emigrate to the US as a first choice rather than returning to their country of origin. In terms of the theories, emigration to the US by Israeli emigrants having high levels of skills fits the neoclassical economics theory in that there are geographic differences in supply and demand for labour force. There are wage differences between the US (relatively high wages) and Israel (relatively low wages compared to the US) which influence workers with relatively high income potentials to emigrate to the US. Apparently, the emigrating family acts as a "rational actor" who examines the emigration options and chooses to emigrate to the US, where the family will receive the greatest net positive returns in exchange for skills. In other words, the two important variables – a large difference between returns in the country of origin and the destination country, and the chances of attaining a remunerative job in the destination country – operate in this case as a mechanism that promotes emigration of those with the highest skill levels, relative to others' skill levels, to the US. 67 In accordance with the assumptions of self-selection theory, it emerged from the current study that the pool of emigrating families is not random. Most of the families emigrating to a particular emigration destination have skills suiting the emigration destination that enable them to have better opportunities in the destination country. Specifically, families that chose to emigrate to the US are families who could also be expected to have relatively high incomes in Israel (due to their educational and/or occupational characteristics), such that there is a strong positive correlation between the income levels of the family in the country of origin and the destination country. In addition, the income distribution in the US is more unequal than that in Israel, which strengthens the US as a desired destination in the selfselection of families that wish to emigrate. This is true also when the emigrant was not born in Israel (as in the case of immigrants from the former USSR): The emigrant will prefer to emigrate to the US in order to maximize returns rather than return to the country of origin, where the returns are lower. From the results of the statistical analyses in this work, one can view the families' choices of the various destinations, overall, as a choice of each family unit according to the skill levels of the couple. Theories that put the family unit at the centre of the choice to emigrate and the migration itself are relevant to the current research because they take into account the complexity of the various characteristics in each family and enable examining both the individual characteristics of each spouse or children as needed. The current research took into account both individual characteristics of each spouse such as education and occupation, as well as characteristics at the couple level such as religion, origin, country of birth, and migration history of at least one spouse, which influence the destination choice of the entire family. Analysis at the family unit level showed that the couple's religion greatly influences the emigration to a particular destination: The variable of the religious composition of the family was highly significant in emigration to all three destinations. In addition, the fact that both spouses were born in Israel, or that at least one of them was born in the emigration destination country, also had a significant influence on the emigration destination, when all 68 other variables were held constant. All of these results demonstrate that the family variables play an important role in characterizing the emigration pattern of the family members, and if one had studied only the individual characteristics of emigrants, without taking into account family characteristics, one might not have noticed the influence of the family characteristics on the emigration, which would prevent taking into account the entire set of causes. Using a global theoretical framework has become meaningful for studies of Israel since the 1990s, when on one hand global companies entered Israel, and on the other hand Israeli companies increased their expansion to external markets. This economic development leads to worker mobility in both directions. However, it mostly encourages educated Israelis with high skill levels to emigrate to the US, due to the fact that many centres of research and development of advanced technological, global companies are located there, and the very large consumer market in the US serves as a desired target for many companies. The theory regarding the role of social networks in maintaining and expanding emigration to particular destinations, seems to operate in emigration patterns of educated Israelis to the US. Social networks that were already formed in Israel, such as in the army, at university, research institutes, and workplaces, promote the emigration of educated Israelis to the US because a social network helps in social and economic integration in the destination country, thus reducing the costs of the emigration and increasing the returns from it. In addition, persons are connected to their colleagues at universities, research institutes, or companies due to the social network phenomenon. Today's modern, vast communication networks enable larger, dynamic social network ties among colleagues in the various work and research fields who live in different countries, thus increasing the possibilities of finding sought-after employment opportunities abroad. The relatively large rate of emigration from Israel to the US among persons holding advanced academic degrees also involves two structural factors that were beyond the scope of the theories. (1) First, the number of academic positions at the Israeli universities is not sufficient to accommodate all of the demand for senior faculty positions. During recent 69 decades, there has been a continual decrease in the number of academic positions in academic institutions per resident, as well as an increase in the number of students per lecturer (Ben-David, 2008). Only a very small number of positions are allocated for young staff members. The policy of reducing faculty size leads to a decrease in the number of fields that the universities can offer to their students and to global research, and an increase in the burden on the remaining faculty members. This load makes it difficult for the researcher to also excel in research, which puts the researcher at a disadvantage compared to researchers in leading Western universities (Podeh, 2006). (2) Second, emigration destination countries prefer migrants with high skills. According to the United Nations (United Nations, 2009b), a growing number of countries prefer migrants with high, needed skills in order to meet a demand for workers in the scientific and technological sectors. Approximately 30 countries, including 17 developed countries, have various policies or programs to promote immigration of highly skilled workers. Since 1965, the US has prioritized the granting of immigration visas to highly skilled workers in order to meet the demand for such workers in the US labour market. The ceiling for visas in this category increased to approximately 140,000 annually according to the immigration law of 1990. In addition, the US has a number of additional visa categories enabling temporary entrance of workers with high skills (United Nations, 2009b). The first reason is a mechanism that pushes academics out of Israel, and the second reason is a mechanism that pulls academics to the US or to other developed countries. The findings of the current research corroborate other research that was quoted in the literature review, mainly, that Israeli families that emigrated to the US are characterized by a high level of skills, and that they constitute the largest portion of emigrants from Israel, both overall and of the group of those with high skills, relative to the other destinations studied. Other researchers came to similar conclusions using different methods. In addition to revalidating their conclusions, the similarity in the conclusions also lends increased validity to the worthwhileness of using the methodology presented in this work. 71 The new database on Israeli families that emigrate from Israel that was created for this work includes various demographic, social, and economic characteristics at both the family and at the individual levels. In addition to enabling reaching the goals of the current research, this series enables additional research with the purpose of characterizing emigration from Israel. One area that is worthy of further study is the family as an emigrating unit. For example, are there differences in the characteristics of various types of families, in terms of different emigration destinations? Are there difference in the characteristics of families that emigrate during different time periods, depending on concurrent events in Israel and in the world? Another area could be focus on emigration countries that were beyond the scope of the current research, such as Canada, Oceania, or the Far East. Research on these topics will broaden the demographic and socio-economic knowledge of this important and interesting population. 71 References Ben-David, D. (2008). Brain drained (CEPR Discussion Paper No. 6717). London, UK: Center for Economic Policy Research. Borjas, G. J. (1987). Self-selection and the earnings of immigrants. American Economic Review, 77(4), 531–553. Borjas, G. J. (1988). Immigration and self-selection (Working Paper No. 2566). Cambridge, MA: NBER Working Paper Series. Borjas, G. J. (1994). The economics of immigration. Journal of Economic Literature, 32, 1667–1717. Castles, S., & Miller, M. J. (2009). The age of migration: International population movements in the modern world. (4th ed.) New York: Palgrave Macmillan. Cohen, Y. (2008). Circular migration in Israel (CARIM research reports 2008/1). Robert Shuman Center for Advanced Studies, San Domenico di Fiesola (FI): European University Institute. Cohen, Y. (2009). Size and selectivity patterns among Israeli-born immigrants in OECD countries (CARIM research reports 2009/12). Robert Schuman Center for Advanced Studies, San Domenico di Fiesole (FI): European University Institute. Cohen, Y., & Haberfeld, Y. (2003). Economic integration among children of Israeli immigrants in the United States. International Migration, 41(4), 141–159. Cohen, Y., & Haberfeld, Y. (2007). Self-selection and earnings assimilation: Immigrants from the former Soviet Union in Israel and the United States. Demography, 44(3), 649–668. Council for Higher Education. (July 2007). Report of the Commission for Examination 72 of the Higher Education System in Israel (headed by Avraham [Beiga] Shohat) [in Hebrew]. Jerusalem: Author. Retrieved from www.che.org.il Eisenbach, Z. (1989). Jewish Emigrants from Israel in the United States. In U. O. Schmelz & S. DellaPergola (Eds.) Papers in Jewish Demography 1985 (pp. 251–266). The Institute of Contemporary Jewry, The Hebrew University: Jerusalem. Gold, S., & Hart, R. (2009). Transnational ties during a time of crisis: Israeli emigration, 2000 to 2004. International Migration, 51(3), 194–216. doi:10.1111/j.1468-2435.2009.00574.x Gould, E. D., & Moav, O. (2007). Israel's brain drain. Israel Economic Review, 5(1), 1–22. Hleihel, A., & Ben-Moshe E. (2002, Sept.). Measuring emigration from Israel. MEDMigr project on migration statistics, workshop on international migration statistics. MEDSTAT programme, Rome. Israel Central Bureau of Statistics (1994). The standard classification of occupations 1994 [in Hebrew]. Tech. Pub. No. 64. Jerusalem: Author. Israel Central Bureau of Statistics. (2009). Society in Israel, 2 [in Hebrew]. Jerusalem: Author. Israel Central Bureau of Statistics. (2010). Statistical abstract of Israel. No. 61, pp. 225, 436 [in Hebrew]. Jerusalem: Author, Massey, D. S., Arango, J., Hugo, G., Kouaouci, A., Pellegrino, A., & Taylor, J. E. (1993). Theories of international migration: A review and appraisal. Population and Development Review, 19(3), 431–466. Michaeli, S., Epstein, A. D., & Kheimets, N. (2007). And the children shall return to within their borders: Toward design of a new policy for reducing emigration and 73 strengthening connections with Israeli citizens residing abroad. In S. Michaeli, A. D. Epstein, & Kheimets (Eds.), National priorities: Immigration and integration in Israel in the beginning of the 21st century (pp. 201–209) [in Hebrew]. Jerusalem: Ministry of Immigrant Absorption – Jewish Agency, and Bar-Ilan University. Ministry of Interior. (2010). Procedure for registering the birth of a child born abroad to an Israeli citizen mother [in Hebrew]. Retrieved from www.moin.gov.il Ouaked, S. 2002. Transatlantic roundtable on high-skilled migration and sending countries issues. International Migration, 40(4), 153–166. Podeh, E. (Dec. 12, 2006). Unsatisfied lecturers [in Hebrew]. Retrieved from www.haaretz.co.il Schellekens, J., & Anson, J. (2007). Introduction: In Israel's destiny: fertility and mortality in a divided society. In J. Schellekens & J. Anson (Eds.), Schnitzer studies in Israeli society (Vol. 12, pp. 1–14). New Brunswick, NJ: Transaction. Sikron, M. (2004). Demography: Israel's population – characteristics and trends [in Hebrew]. Jerusalem: Carmel. Sikuler, N., & Kurz, A. (Jan. 21, 2010). Calcalist panel: They don't flee because of the salaries [in Hebrew]. Retrieved from www.calcalist.co.il Skeldon, R. (2009). Of skilled migration, brain drains and policy responses. International Migration, 47(4), 3–29. Stark, O., & Bloom, D. E. (1985). The new economics of labor migration. American Economic Review, 75, 173–178. Todaro, M. P. (1969). A model of labor migration and urban unemployment in less developed countries, American Economic Review, 59(1), 138–148. Todaro, M. P., & Maruszko, L. (1987). Illegal migration and US immigration reform: 74 A conceptual framework. Population and Development Review, 13(1), 101– 114. Toren, N. (1980). Return to Zion: Characteristics and motivations of returning emigrants. In E. Krausz, (Ed.). Studies of Israeli Society: Vol. I, Migration, ethnicity and community. New Brunswick, NJ: Transaction. UNHCR – The UN Refugee Agency. (2012). Figures at a glance. New York, NY: Author. Retrieved from www.unhcr.org United Nations. (2009a) Guidelines on the use and dissemination of data on international immigration to facilitate their use to improve emigration data of sending countries (note by the United Nations Economic Commission for Europe/Eurostat task force on measuring emigration using data collected by the receiving country). New York, NY: Author. United Nations, Department of Economic and Social Affairs, Population Division. (2009b). International migration report 2006: A global assessment. New York, NY: Author. Retrieved from www.un.org/esa/population/publications /2006_MigrationRep /report.htm 75 6. Appendix A: Method of Constructing the New Database of Israeli Families Abroad The research population included male and female, current or past residents of Israel, included in the Population Register with their children through age 17, in cases in which there was an indication in the Register that at least one child was born abroad, during 1996– 2008. This research population comprised the database of the Israeli families abroad. The database was constructed and reviewed in three major stages: (a) creating the basis for the database, (b) adding characteristics to the series from CBS files with identifying information, and (c) reviewing whether the records from the database were found in other CBS "possible emigrants" databases. Following is a diagram and explanations of the three stages: 76 (A) Creating the database Diagram of the Stages of Creating the Database CRA IC File 1: Source: Children born abroad during 1996–2008 to an Israeli citizen, and whose status is "CRA". Source: Population Register 04/2009. Number of records: 27,480 File 2: Source: Children born abroad during 1996–2008 to an Israeli citizen, and who immigrated to Israel. Source: Multi-year data file of ICs, based on the Population Register, extracted from the CBS's migration system Number of records: 25,565 File 1b: Constructing administrative families: Adding parents and children through age 17 to the records in File 1. Source: Population Register 04/2009. Number of records: 70,035 File 2b: Constructing administrative families: Adding parents and children through age 17 to the records in File 2. Source: Population Register 04/2009. Number of records: 67,547 File 3: Merging of Files 1b and 2b. Attaching all characteristics from the Population Register 04/2009 to each record. Number of records: 133,539. This file included the entire research group. The variable "country of birth" of the records in File 1 + 2 was copied to the other family members and labelled "destination country". 77 In the first step, relevant records were identified. These were records with Register status implying "CRA" in the Population Register 04/2009 (the determinant date for this research was December 31, 2008) and records with the status "IC" in the multi-year file of ICs. This resulted in 49,045 records that formed the basis for the database. In the second step, the family of every record, including (at least) one parent or two parents and siblings up to age 17, was constructed from the Register. This age was chosen as an age prior to legal independence; moreover, at this age, the vast majority of children are still under their parents' care and will migrate together with them. The family was constructed according to an algorithm for building nuclear families that was used to build administrative families in the 2008 Population Census, with one difference: In contrast to the Census guidelines, in this research file the geographic location of the various family members did not serve as a criterion for linking the records. Below is a brief description of the algorithm for building nuclear families: 78 Step 1: Pairing children with one or two parents: Research group Marking of records of persons who are unmarried, have no children, through age 17 (i.e., the records in File 1 and File 2) Was a mother located? No Yes Was a father located? Joining mother to her children Joining mother to her children No End of process. No family found for the person. Yes End of process. Attaching father's ID number and that of his spouse and her children to the child's record. Was a spouse of the mother located? No End of process. Family created of mother and her children. Yes End of process. Attaching mother's ID number and her children to the records of her spouse and his children. 79 Step 2: Attempt at locating a spouse for parents who remained without spouses after Step 1 Marking in the database of all persons without spouses after Step 1 Was a lone male partner (with or without children) located for lone women (with no children according to the definition of the model)? No No family was found Yes Pairing of the husband and his children with the record of the wife After application of this algorithm and screening out of surplus records, 133,509 records remained in the base file of the database. These constituted 34,057 families who were assumed to be currently living overseas or lived there for a substantial period of time in the past during the research period (1996–2008). After merging this file with the Population Register file, unneeded variables and duplicate records were removed, and a unique number was assigned to all family members. The same emigration destination was assigned to each individual in the family according to the country of birth of the child born abroad. If the family had more than one child abroad, and the children were born in different countries, the highest country code was chosen, which gave slightly more priority to the US (code 710), Canada (code 700), or Australia (code 860. These countries accounted for more than 53% of the emigrants in the file. 81 Following are the 10 emigration destination countries with the largest number of persons, in descending order: Emigration destination country Total Number of persons in research file 133,502 Thereof: US 60,169 France 10,559 United Kingdom 9,230 Canada 7,220 Russian Federation 4,921 Germany 4,633 Australia 3,078 Belgium 3,051 Netherlands 2,287 Ukraine 2,174 Then, a variable was constructed for relationship to the family representative, according to the Population Register. The oldest person in the family, man or woman, was assigned Code 1, the spouse of that person (if married to a spouse in the Register) was assigned Code 2, and the children of the couple or of the single parent were assigned Code 3. After building this variable, the database had 34,046 records with Code 1, 22,553 with Code 2, and 76,903 with Code 3. The variable "marital status" was handled as follows: For married persons, if there was an "ID number of spouse" for only one of the spouses, the other one was completed accordingly. If the "ID number of spouse" of the person assigned Code 1 was not found in the Population Register, the spouse was deleted (this was attributed to an error in 81 recording). For divorced or widowed persons, if there was an "ID number of spouse", it was deleted. After these corrections, the marital status data of the parents in the database was as follows: Absolute numbers Marital status Percentages out of known cases Gender Total Men Gender Women Total Men Women Total 56,376 26,648 29,728 100.0 100.0 100.0 Single 2,691 838 1,853 4.8 3.1 6.2 Married 51,569 25,429 26,140 91.5 95.4 87.9 Divorced 1,957 337 1,620 3.5 1.3 5.4 Widowed 159 44 115 0.3 0.2 0.4 Unknown 223 (B) Adding Characteristics to the Research Group from Identified CBS Files In the second stage of building the database of Israeli families, additional identified CBS files were added to the database. Identified files are data files containing individual records with ID numbers. These files enriched the scope of data about the persons in the research group. Selected characteristics from five identified files were added to the base file: (a) the Population Register 4/2009, (b) the 20% Sample File from the Population and Housing Census 1995, (c) the database of degree recipients 1983–2008, (d) the multi-year file of immigrants to Israel 1984–2008, and (e) the DMPIR during 1996–2008. 82 Diagram of Adding Characteristics to the Database from Additional Files File 3, Database: 133,509 records; 34,057 families Database of degree recipients 1983–2008. The DMPIR during 1996– 2008. Added educational characteristics: type of degree, year of degree, and field of study of the degree. Adding characteristics of record status and year of emigration from Israel. The following were linked with this file, with the following statuses: Immigrants (27,330 records linked), family unification (2,457 records linked), returning residents (12,952 records linked), exiting emigrants who did not yet return (23,929 records linked). 15,188 records of persons aged 18 and over were linked (by ID number). Population Register 04/2009. Added demographic characteristics: sex, age, marital status, religion, country of birth, nationality, Population Register status, year of immigration, year of entry to Israel, year of birth, and year marital status took effect. All records in File 3 were linked to the Register; 99,954 were found to have "active" status and 33,555 had "CRA" status. Multi-year file of immigrants to Israel 1984– 2008. Includes immigrants who immigrated to Israel during these years. Added the characteristic of last country of residence before immigration. 20% Sample File from Population Census 1995. Added the characteristics education, occupation, marital status at the time of the census. 6,400 records linked 13,947 records linked. At the end of Stage 2, the expanded database included 55 variables. Some were taken as is from the files described above, and others were constructed especially for this research based on the data in these source files. The variables that were added to the file were varied, and enable research in many fields of demographic and socio-economic statistics. The database is in SAS format. The following is a description of the structure of the research 83 file, which is the new database of Israeli families abroad that was constructed in this research: Structure of the Database File No. Variable Name Values Notes 1 ID Number Source: Register 04/2009 2 Father's ID Number Source: Register 04/2009 3 Spouse's ID Number Source: Register 04/2009 4 Mother's ID Number Source: Register 04/2009 5 ID Number of Exists for married couples who are Married Spouse both in the research file and in the Population Register. The variable was not relevant for married persons without a spouse in the Register, and for divorced and widowed persons. Source: Register 04/2009 6 Destination country Country codes of Ministry of Interior Calculated variable. Derived from the of emigration from model. Country of birth of the Israel records in the source files copied to the variable "destination country" for all family members. 7 Code for source file 1 = CRA file of persons born abroad of record Calculated variable. Created as part of the procedure of creating the research group. Sources: Population Register and IC files 2 = IC file of persons born abroad 8 Source of record in 1 = from source files: File 1 (CRA) or Calculated variable. Created as part procedure of creating File 2 (IC) of the procedure of creating the administrative family administrative family in the model. 84 No. Variable Name Values Notes 2 = family members of records from source files who were added in procedure of creating administrative family 9 Unique number for Calculated variable. Created as part administrative family of the procedure of creating the administrative family in the model. 10 11 Number of persons 1...17 Calculated variable. Created as part in administrative of the procedure of creating the family administrative family in the model. First person in family 1 Calculated variable. Determined by the earliest birth date in the family. empty = all other family members 12 Spouse exists in the 1 = Married, person of first record Calculated variable. Register (one of the parents) Source: Register 04/2009 2 = Married, spouse in research file Variable is relevant only for married of first person persons. 3 = Married, no spouse in research file 13 Family relationship to 7 = First in family record Calculated variable. Value 2 exists if first record in married and spouse has an ID research group number of the research file. model 2 = spouse Source: Register 04/2009 3 = child 14 Age 0...80 Reference date: Dec. 31, 2008. Source: Register 04/2009 85 No. Variable Name Values Notes 7. 1 = Jew Source: Register 04/2009 Grouped religion 2 = Moslem 3 = Arab Christian 4 = Other Christian 5 = Druze 6 = Unclassified by religion 16 Country of birth from Country codes of Ministry of Interior Source: Register 04/2009 Citizenship code Citizenship codes of Ministry of Source: Register 04/2009 from the Register Interior Nationality code from Nationality codes of Ministry of the Register Interior Marital status code 10 = single the Register 17 18 19 Source: Register 04/2009 Source: Register 04/2009 from the Register 20 = married 30 = divorced 40 = widowed 20 Sex code from the 1 = male Source: Register 04/2009 2 = female Source: Register 04/2009 According to the 2008 list of localities Source: Register 04/2009 Register 21 Locality code of Ministry of Interior 22 Year of immigration Source: Register 04/2009 23 Year of first entry to Source: Register 04/2009 Israel 24 Year of last entry to Source: Register 04/2009 Israel 86 No. Variable Name 25 Values Year of immigrant Notes Source: Register 04/2009 entry 26 Year of birth Source: Register 04/2009 27 Year marital status Source: Register 04/2009 took effect 28 Year of last exit 29 Code of status in the Source: Register 04/2009 1 = active Register "CRA" includes other similar statuses: 3, 5, 54, 61, 71, 72. Source: Register 04/2009 51 = CRA 9. Included in 1 = the record has status of family Calculated variable. Source: DMPIR population through unification in the DMPIR 2008 family unification in the DMPIR 2 = the record does not have status of family unification in the DMPIR 31 Included in 1 = the record has status of Calculated variable. Source: DMPIR population through immigrant in the DMPIR 2008 immigration in the DMPIR 2 = the record does not have status of immigrant in the DMPIR 32 Code of entry status 5 = immigrant Source: DMPIR 2008 in the DMPIR 8 = family unification 9 = returning residents 33 Code of exit status in 7 = emigrant Source: DMPIR 2008 the DMPIR 87 No. Variable Name 34 Values Notes Year of emigration Separate, independent datum for from Israel – DMPIR each year. A datum existing for one ("emigration") family member was copied to the other family members. In the case of two or more different data, the earliest year was used. Source: DMPIR 2008 35 Year of emigration Created from the merger of stock from Israel – stock of files from 1995 through 2008 and emigrants CRA through 2003. The stock ("emigration") consists of all "possible emigrants" each year since 1995. There is data for 2004–2009. A datum existing for one family member was copied to the other family members. In the case of two or more different data, the earliest year was used. 36 Source of the record 1 = model within the database Source: database of potential emigrants from 2008 Census of potential emigrants in emigrant model of Census 2008 3 = earlier database of potential emigrants known as "Dan 95" 4 = citizenship code 23 5 = National Insurance records 88 No. Variable Name 37 Values Notes Database of potential 1 = the record exists in the database Calculated variable. Source: emigrants from 2008 of potential emigrants from 2008 database of potential emigrants from Census Census 2008 Census 9 = the record does not exist in the database of potential emigrants from 2008 Census 38 Left population as an 1 = the record exists with status of Calculated variable. Source: DMPIR emigrant in the emigrant in the DMPIR 2008 2 = the record does not exist with Applies to persons who were given status of emigrant in the DMPIR emigrant status at the end of 2008 Entered population 1 = the record exists with status of Calculated variable. Source: DMPIR as a returning returning emigrant in the DMPIR 2008 DMPIR 39 resident in the DMPIR 2 = the record does not exist with status of returning resident in the DMPIR 40 Aged 15 or over in 0 = aged 0 through 14 1995 Census Source: 20% Sample from the 1995 Census 1 = aged 15 or over 89 No. Variable Name 41 Family relationship to Values Notes empty = unknown Source: 20% Sample from the 1995 first record in Census household in 1995 Census 1 = first record 2 = spouse 3 = child 4 = parent 5 = sibling 6 = brother- or sister-in-law 7 = son-in-law or daughter-in-law 8 = mother-in-law or father-in-law 9 = grandchild 10 = other relative 11 = non-relative 12 = nephew or niece 15 = great-grandchild 16 = grandparent 42 Total years of 0 = did not study at all Imputed variable. Does not include schooling in 1995 years of study in yeshiva (religious Census academy). Source: 20% Sample file from the 1995 Census xx = total years of schooling 98 = children under age 15 43 Record exists in 20% 1 = the record exists in 20% Sample Calculated variable. Source: 20% Sample from the file from the 1995 Census Sample file from the 1995 Census 1995 Census 2 = the record does not exist in 20% Sample file from the 1995 Census 91 No. Variable Name 44 Occupation: first digit Values Notes empty = not relevant Alphanumeric variable. 0 = academic professionals Source: 20% Sample from the 1995 in 1995 Census Census 1 = associate professionals and technicians 2 = managers 3 = clerical workers 4 = agents, sales workers, and service workers 5 = skilled agricultural workers 6–8 = industry, construction, and other skilled workers 9 = unskilled workers X = under age 15 Y = unknown occupation + other 45 Marital status in 1995 1 = married Census Source: 20% Sample from the 1995 Census 2 = divorced 3 = widowed 4 = single 46 47 Unique ID for Created during the editing process in household in 1995 the Census. Source: 20% Sample Census file from the 1995 Census Participation in 0 = not in annual civilian labour force annual civilian labour Source: 20% Sample file from the 1995 Census force? 1= in annual civilian labour force 8 = children under age 15 91 No. Variable Name Values Notes 73 Type of record in 1 = Person enumerated in regular Source: 20% Sample file from the 1995 Census household 1995 Census 2 = Person enumerated in regular household outside of a locality 3 = Person enumerated in a kibbutz household 4 = Person enumerated as a resident of an institution 5 = Person enumerated in a household in an institution 49 Highest degree in 0 = did not study at all 1995 Census Imputed variable. Source: 20% Sample from the 1995 Census. For persons aged 15 and over. 1 = no diploma or certificate 2 = completed elementary or lower secondary school 3 = completed upper secondary school 4 = matriculation certificate 5 = non-academic post-secondary certificate 6 = first degree 7 = second degree 8 = third degree 9 = other certificate 50 Field of academic 0 = general humanities study For persons aged 18 and over. Source: database of degree recipients 1983–2008 1 = languages, literature, and 92 No. Variable Name Values Notes regional studies 2 = education and teacher training 3 = art, crafts, and applied arts 4 = social sciences 5 = business and management science 6 = law 7 = medicine 8 = paramedical studies 9 = mathematics, statistics, and computer science 10 = physical sciences 11 = biological sciences 12 = agriculture 13 = engineering and architecture 14 = special programs, misc. 51 Highest academic 10 = first degree degree For persons aged 18 and over. Source: database of degree recipients 1983–2008 20 = second degree 30 = M.D. (second degree in medicine) 40 = third degree 50 = teaching certificate 60–69 = non-academic certificate 52 Year of receipt of academic degree 7337–9..3 For persons aged 18 and over. Source: database of degree recipients 1983–2008 93 No. Variable Name 53 Last country of Values Notes Country codes of Ministry of Interior For persons who previously had residence before immigrant or potential immigrant immigrating to Israel status, during 1984–2008. Source: Multi-year file of immigrants. 54 Database of stock of 1 = found in stock of emigrants in Calculated variable. Built from emigrants – DMPIR DMPIR combining stock files of 1995 1995–2008 through 2008 and the CRA file through 2003. The stock consists of all "possible emigrants" each year since 1995. 55 "Surplus" records 1= Record exists in "surplus" records Calculated variable. Basis created that were in the based on the 1995 Census. Possibly Population Register long-term residents abroad. in 1995 but were not found in the field at the time of the census enumeration (C) Checking for Database Records in Other "Possible Emigrant" CBS Databases In the third stage, the records in the database were compared with the various "possible emigrant" databases existing at the CBS. It should be noted that the purpose of the current research was not to locate and estimate the number of Israelis who live abroad permanently, but rather to locate Israeli families spending a substantial period of time abroad and examine their characteristics by the various emigration destination countries. Thus, there are families in the research file that were located as individuals who emigrated from Israel and had not returned by 2008, or families that returned during the research period, and are defined as "returning residents" or "immigrant citizens" (IC). At the same time, in order to examine the methodology that was developed, it is interesting to see where it stands compared to the 94 other data sources of possible emigrants. In future research, it will be possible to focus only on the records that appear both in the database and in a particular data source of possible emigrants in order to gain validation from different sources on emigrants living outside of Israel. Diagram of Comparing the Database Records with "Possible Emigrant" Databases at the CBS File 3, Database: 133,509 records Database of Israeli residents abroad, 2008 Census. Before the 2008 Census, a database of possible emigrants was created based on data from various files, such as border control, National Insurance, and individuals who were not enumerated in two previous censuses. Results: 40,054 records from File 3 were found in this database. DMPIR. Location of records from File 3 in a combined file of all ID numbers appearing at least once in the database ("stock") of emigrants of the DMPIR during 1995–2008, for each year. Results: 26,678 records from File 3 were found in this database. A "surplus" database from the 1995 Population Census, which began as records that were in the Register in 1995 but were not found in the field at the time of the census enumeration and were not linked to the population of the 1995 Census, and which constitute a component of the DMPIR since 1995. The year 1995 was selected for checking because the database gradually emptied since the Census. Results: 14,457 records from File 3 were found in this database. In these three CBS databases of possible emigrants, together, 81,189 records were found from File 3. These records constitute 61% of all records in the database. There were 13,687 records common to the database of emigrants of the DMPIR and the database of Israelis abroad from the 2008 Census. Records from File B1 (CRA) in the "possible emigrant" databases: 95 In the database ("stock") of emigrants of the DMPIR, 14,102 records originating in File B1 were found. In the "surplus" database from the 1995 Population Census, which began as records that were in the Register in 1995 but were not found in the field at the time of the census enumeration and were not linked to the population of the 1995 Census, 9,609 records originating in File B1 were found. In the database of Israeli residents abroad, 2008 Census, 29,233 records originating in File B1 were found. Total: In the "possible emigrant" databases, 52,944 records were found, constituting 76% of the total records in File B1. Records from File B2 (IC) in the "possible emigrant" databases: In the database ("stock") of emigrants of the DMPIR, 12,576 records originating in File B2 were found. In the "surplus" database from the 1995 Population Census, which began as records that were in the Register in 1995 but were not found in the field at the time of the census enumeration and were not linked to the population of the 1995 Census, 4,848 records originating in File B2 were found. In the database of Israeli residents abroad, 2008 Census, 10,821 records originating in File B2 were found. Total: In the "possible emigrant" databases, 28,245 records were found, constituting 42% of the total records in File B2. 96 7. Appendix B: Reliability of the Database as a Proxy for the Population of Israeli Families Abroad Although every research study on a particular population has limitations and some deficiency, and there is no study that fully covers the whole population, the main methodological contribution of the current research was the creation of a new, diverse database of Israeli families abroad that enabled broadening our knowledge of the population of these families, according to definitions and variables used by the CBS and in Israel, while using the variable of emigration destination country of the family. This was accomplished within the framework that the research population included families that emigrated during the investigation period, to whom children were born abroad, and which registered their children at the Israeli consulate abroad. Thus, the question arises: Can the research population included in the new database be used as a proxy (i.e., a reliable representation) for the rest of the population of Israeli emigrant families? To address this question, the reliability of the methodology used in the current research was examined. Following are the results of the examination. Israeli society is characterized by high marriage and fertility rates relative to the average in the other developed countries (as well as in comparison to the OECD countries). Approximately 70% of Israeli households in 2002 consisted of a married couple, with or without children, and a household had an average of 3.36 persons – larger than in the developed countries (Central Bureau of Statistics, 2009; Schellekens & Anson, 2007; Sikron, 2004). Based on these characteristics, it can be assumed that the emigrants from Israel are also characterized by a large percentage of families and parents with children, even if they were not married or did not have children when they left Israel. It is reasonable to assume that behavioural norms they acquired while growing up in Israel influence them also as adults, and that most marry and have children. Recall that the current research did not attempt to draw conclusions about all emigrants from Israel, but rather focused on Israeli 97 families that live abroad or lived abroad in the past for a substantial period of time, as per the definitions of the research population. Registration of Children Born to Israeli Citizens Abroad in the Population Register When a child is born to an Israeli citizen abroad, and the parents wish to register the child in the Israeli Population Register, they must fill out a form called "Notification of Birth of an Israeli Citizen Abroad" (MR/23). Then, the Ministry of Interior carries out the "Procedure for Registration of a Child Born Abroad to an Israeli Citizen Mother" (Procedure No. 2.2.0004; Ministry of Interior, 2010). This procedure is carried out at an Israeli consulate abroad, or, if in Israel, at an office of the Ministry of Interior. Under this procedure, after verifying one or both parents' personal information (even if they are not married to each other), as well as that of the child, a new record is recorded in the Population Register. If the parent declares that he or she is not settled in Israel, the record is immediately assigned a status of "citizen residing abroad" ("CRA"). In the current project, these families were combined into a file entitled CRA. Later, if the person titled CRA enters Israel with the intent of setting, his status is altered to "immigrant citizen" ("IC")(Central Bureau of Statistics, 2010, p. 225). The children born abroad are also an indication that their families lived outside of Israel for a substantial period of time, before the child returned to Israel without his or her family. In the current research, these families were aggregated into a file entitled IC. The two files CRA and IC constituted the basis for the new database of Israeli families abroad (for more information on the process of constructing these files, see Appendix A). Thus, the families included in the database have at least one parent who was an Israeli citizen and resident in the past. Because Israeli society is characterized by relatively high family cohesiveness, it is reasonable to assume that many families that emigrate from Israel still maintain connections with relatives in Israel, including physical visits to Israel. By law, when entering and exiting Israeli borders, Israeli passports must be used for all of the family members, which encourages the parents to register the children at an Israeli consulate abroad prior to visiting Israel. In addition, persons who emigrate via means such as repeat or cyclical migration, 98 which have become more common in recent years, or who live alternately in two countries, tend to leave open the option of returning to Israel. It is therefore very likely that they have a motivation to register their children's births abroad at an Israeli consulate there. These reasons increase the likelihood that a substantial portion of the population that emigrates from Israel registers their children who are born abroad, even if they are settled in another country at that stage in their lives. Comparison Between the Population Included In the Database and the Parallel Population in the US Population Census The US was chosen as the comparison group because it is home to the largest group of Israelis abroad, and because there are US data available for research in the form of population files from the US Census that can be downloaded from the Internet. Data on families of Israelis in the US were produced from the US Population Census file for 2000, which included a sample of 1% of the residents of each state within the US borders. To match the two comparison files, a group of families was derived from the US Population Census file that met the following criteria: at least one parent was born in Israel, the family has children under age 18, the family consists of only parents and children, and at least one of the children was born in the US (N = 149,498). From the current Israeli database, in 2008, a group of families was derived that met the same criteria: families that emigrated to the US in which at least one parent was born in Israel, the family has children under age 18, the family consists of only parents and children (this was derived from the way the file was created), and at least one of the children was born in the US (N = 51,272). The results of the comparison between the two groups on a number of selected indicators is presented below. A) Average family size: In the database of the current research: 4.1 persons; in the US Census: 4.6 persons. B) Comparison of age composition: As can be seen in the age pyramid charts, the age structures of males and females in the two groups are very similar, with the exception of small variations in a few age groups. 99 Age Pyramid Comparison Between Data of Israeli Families Abroad and a Parallel Group from the US Census: Males Age Pyramid Comparison Between Data of Israeli Families Abroad and a Parallel Group from the US Census: Females 111 C) An additional comparison was made between the two groups regarding education. The number of holders of academic degrees was examined, and based on this, holders of first degrees, and holders of second degrees and above. Again, as can be seen from the diagram below, the results indicated great similarity between the groups: Percentage US Census 2000 100.0 90.0 80.0 70.0 60.0 50.0 40.0 30.0 20.0 10.0 0.0 37.7 Series – Israeli Families Abroad 2008 33.2 23.8 21.4 13.9 11.8 Total academic degrees First degree* Second degree and above* * Calculated from total holders of academic degrees The slightly larger percentage of holders of academic degrees in the data from the US Census might be explained by the fact that a not inconsiderable number of emigrants move to the US to further their academic studies in the various frameworks there, thus slightly raising their average educational level above what it was when they left Israel. Comparison Between the Research Group and the Register of Doctoral and Research Students Abroad A comparison was made with another outside source of information. The report of the Commission for Examination of the Higher Education System in Israel, known as "the Shohat Commission" (Council for Higher Education, 2007), contained an appendix (Appendix 2) that was a summary of the results of a register of young Israeli doctoral and research students who reside abroad. The summary included approximately 900 men and women. In the current research, a comparison was made between the data of holders of 111 doctoral degrees in the major fields of study as recorded in the Shohat report in 2007 (Council for Higher Education, 2007), and the academic fields of study in which persons received doctoral degrees (PhDs) in Israel among the database of Israeli families abroad in the current research (source of data: CBS database of persons receiving academic degrees, 2008). Note that it makes sense to compare these two sets of data in that 73% of the holders of doctoral degrees in the Shohat Report (Council for Higher Education, 2007) received them in Israel. Holders of doctoral Holders of doctoral degrees Major field of research degrees in Shohat in current database Report Physical sciences/biological sciences 45% 55% Computer sciences/mathematics 12% 13% administration/Management/Economics 11% 8% Humanities 6% 2% Medicine 6% 5% Social sciences 6% 7% Engineering 3% 8%* N = 531 N = 1,171 Business *Including architecture. The percentages of holders of doctoral degrees are quite similar, taking into account the limitations of matching the classifications of fields of study between the two groups. The classification groups are slightly different in the two sources. In addition, the CBS database of degree recipients indicates the field of study in which the doctorate was received, whereas the Shohat Report (Council for Higher Education, 2007) indicates the current major field of research in the US. There might be slight changes in the field in which a person is involved between these two points in time. 112 To sum up, from comparison between the database of families of Israelis abroad and the external sources of information that were examined, it emerges that the data on the families in the database are reasonably similar to data obtained from various external sources. This fact increases the reliability level of the findings and the validity of the methodology that was developed as suitable for the research population. 113 8. Appendix C: Description of the Variables in the Research File for Testing the Research Hypotheses (Married Couples with Measurable Skills) Variable Name Source of Data Values Notes Dependent variable (explained variable) Grouped emigration destination countries Variable calculated from Population Register 1 = US Calculated variable derived from the model. Variable shared by both spouses. 2 = Central and Western Europe 3 = Former USSR and Eastern Europe 4 = Other Independent variables (explanatory variables) Demographic variables Gender Age Religion: Both spouses Jewish Population Register Population Register Population Register Country of birth Population Register Both spouses born in Israel Variable calculated based on Register Country of birth same as emigration destination country Variable calculated based on Register Immigrant Population Register Male/female 7 = both spouses Jewish 0 = at least one spouse not Jewish Country codes of Ministry of Interior 7 = both spouses born in Israel 0 = at least one spouse not born in Israel 7 = at least one spouse born in, and then emigrated to that country 0 = neither spouse born in, and emigrated to that country 7 = Had immigrant status in the past 0 = Did not have immigrant status in the past Variable per individual Variable per individual Calculated variable Variable per individual Variable common to both spouses. Variable common to both spouses. Based on immigration year in the Register. Variable per individual. Socio-economic variables Education - First degree (B.A.) Education - Second degree (M.A.) CBS database of degree recipients or highest degree in 1995 Census CBS database of degree recipients or highest degree in 1995 Census 1 = has first degree 0 = does not have first degree Variable per individual 1 = has second degree 0 = does not have second degree Variable per individual 114 Education - M.D. (second degree in medicine) Education - Third degree (PhD) Education – Second degree or higher Other education (less than academic) "High scientific" academic field Occupation requiring a high skill level CBS database of degree recipients CBS database of degree recipients or highest degree in 1995 Census Calculated variable based on CBS database of degree recipients or highest degree in 1995 Census Calculated variable based on CBS database of degree recipients or highest degree in 1995 Census Calculated variable based on CBS database of degree recipients Variable calculated based on 20% Sample from the 1995 Population Census 1 = has M.D. 0 = does not have M.D. 1 = has third degree 0 = does not have third degree Variable per individual 1 = has second degree or higher 0 = does not have second degree or higher Variable per individual 7= has other education 0 = does not have other education Variable per individual. Includes "did not study at all" and all other types of non-academic certificates A field of academic study that requires a person to have a high level of skills to be accepted and complete the degree. The field can be more easily adapted to the international labour market. Includes medicine, mathematics, statistics, computer science, physical sciences, biological sciences, engineering, architecture. Variable per individual 7 = has a degree in a high scientific academic field 0 = does not have a degree in a high scientific academic field 7 = Has occupation requiring a high skill level 0 = Does not have occupation requiring a high skill level Does not include doctorate in medicine. Variable per individual Includes these classifications: 0 = Persons with academic professionals occupation 1 = Persons with associate professionals and technicians occupation. According to the CBS Standard Classification of Occupations 1994. Variable per individual 115 116 9. Appendix D: Selected Distributions from the Database of Israeli Families Abroad, 1996–2008 Selected Data Total 10 Countries Chosen Emigration Destination Country Russian GerAusBelNetherUS France UK Federmany tralia gium lands ation Total 113,836 ..67.3 7.6..3 3699. 6699. 76397 76.99 96.63 96..7 96936 % percentages Religion - % of total emigrants to destination country Jews 7.9697. 37.7 3..6 37.7 33.3 6..7 .3.6 39.. 37.9 33.. Arabs 76663 7.7 ..3 7.3 ..9 7.9 73.9 7.9 7.. 9.6 Others* .6677 7.. 9.7 ..6 ..3 93.. 77.7 9.3 7.6 6.6 Marital status - % of persons aged 15 and over who emigrated to destination country Single 36399 77.3 99.9 7..9 76.7 7... 7..3 77.3 73.7 9..3 Married 7769.7 39.. 67.9 39.7 3..9 66.. 63.. 39.6 63.. 67.9 Divorced 76..6 9.7 9.9 9.7 9.9 ..7 7.3 9.7 9.9 7.3 Widowed 79. ..9 ..7 ..7 ... ..9 ..9 ..7 ..7 ..7 Number of persons in household - % of total emigrants to destination country 7-9 36... ..9 6.6 ..6 ..7 73.. 73.3 6.. 7.. 73.. 9-7 .76339 76.7 93.. 76.7 .9.. ...9 ...9 73.. 93.9 .7.3 .-. 936399 93.7 9... 97.. 99.7 7..6 99.. 9... 79.3 97.. 7+ 7.6.99 3.9 7..3 77.3 3.. ... 9.3 3.3 77.9 9.9 Country of birth - % of total parents in file who emigrated to destination country Israel 9.63.9 67.7 93.. ...9 ...3 76.7 .3.6 .3.6 ...3 69.6 Born in and emigrated to same 669.7 7..3 79.3 76.7 ..6 93.3 3.. 77.9 77.. 7..6 destination country Population Census 1995 - % of emigrants aged 15 and over who emigrated to destination country 13 years and over of 967.7 .9.7 93.3 .9.3 .9.. .7.9 76.7 ...7 76.3 79.7 schooling Highest degree: Second 97. ..3 9.. 7.3 79.. 7..6 3.6 ... 7.7 ... degree or higher CBS degree recipients database - % of persons aged 15 and over who emigrated to destination country Highest degree: Second 76777 97.6 96.7 93.7 99.3 93.7 97.7 9..3 97.. 96.. degree or higher Field of study for degree: Humanities, languages, 9637. 73.9 9..3 97.. 9... 93.. 93.3 99.6 96.3 9..6 literature, art Social sciences, business and .693. 93.3 7... 77.3 97.9 77.9 99.3 99.3 7..9 77.. management science, law Medicine 796 9.. ..3 7.. ..3 7.7 7.. 9.7 ..7 9.9 Absolute numbers Canada Uk-raine 96767 .... 7.7 79.9 77.. 63.7 ..3 ... 77.3 .3.7 7..7 ..9 9... 93.. ...7 77.9 117 Mathematics, statistics, computer science Physics, biology 76939 7... ..3 ..3 3.. 3.. ..9 79.. ..9 ..9 769.7 7..7 3.9 6.. 6.. ..7 77.7 6.3 9.3 ..3 * Non-Arab Christians, members of other religions, and not classified by religion in the Population Register. 118
© Copyright 2026 Paperzz