Keywords

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

4.1 Introduction

Entering the labour market, gaining adequate employment and pursuing a successful occupational career are central events in the life course of each individual. Among immigrants such events might occur several times in various societal contexts, implying that opportunities and constraints of these specific social circumstances have to be taken into account. Migration in itself is interconnected with the above-mentioned life course events, so that understanding these interrelationships and figuring out how these might vary across various immigrant origins and socio-economic groups are important aims of the migration-related life course research.

The human capital model is considered to be the dominant paradigm in immigrant labour market integration research (Chiswick 1978, 1979, 2005; Kalter 2003). According to its argumentation, immigrants with higher levels of human capital, above all education and labour market experience, are more likely to be quickly and successfully integrated into the receiving society. Empirical evidence largely supports this idea, but recent studies have also shown that these assumptions cannot, by far, account for all group and all country differences (Kalter and Granato 2007; Kogan 2011). While in some countries educational credentials are extremely predictive of immigrants’ labour market success, in other countries additional mechanisms seem to mediate the effect of education on the newcomers’ fortunes to a large degree.

In modern economies, even highly-educated immigrants need additional resources to find adequate employment. Higher qualifications are often valued by employers only if they can be supported by fluency or even eloquence in the host country’s language (Esser 2006; Kossoudji 1988). Immigrants also have to possess specific knowledge of how the labour market of the host society functions. Furthermore, access to good employment opportunities is often facilitated by the availability of social contacts and network resources (Granovetter 1973; Portes 1995; Lin 1999; Portes and Rumbaut 2001). Finally, characteristics that are not captured by formal qualifications, like motivation or readiness to take a risk, might also prove to be relevant for the productivity of immigrants and thus for their successful integration into the economic structures of the receiving society (Chiswick 1978; Borjas 1990; Cohen and Haberfeld 2007). The relative importance of educational credentials and of each of these additional factors depends heavily on the institutional settings of the receiving society, like legal restrictions, labour market regulations etc. Thus the labour market integration of immigrant groups in receiving countries can only be understood by accounting for the interplay between immigrants’ individual resources and host countries’ structural constraints.

These ideas are well known in integration theory, and an increasing amount of empirical studies has proved the importance of the one or the other aspect. To date, however, hardly any empirical studies regard all these theoretical arguments holistically and test them simultaneously with adequate data and an adequate design. In this paper we intend to do so by means of a comparative analysis of the labour integration of FSU immigrants in Germany and Israel. We make use of primary data that were collected specifically to test the above-mentioned mechanisms.

We start by presenting the general theoretical arguments on the role of individual resources for immigrants’ labour market success in more detail, and continue by explaining the rationale behind a comparison of FSU immigrants in Germany and in Israel. Then we describe our data and the variables that capture the theoretical constructs. After this, we present our main results. We conclude with a short summary and a discussion of our findings in the light of the study’s major research aims.

4.2 On the Role of Individual Resources for Immigrants’ Labour Market Success

Labour markets represent arenas in which workers exchange their labour in return for wages, status and other benefits (Kalleberg and Sørensen 1979: 351). This process is two-sided. On the demand side of the labour market, employers seek to maximise their profits, sometimes not free of individual preferences, and thus determine the set of opportunities for prospective employees. On the supply side, job seekers try to maximise their income and other labour market outcomes by investing their available resources, which mainly determines their choice of opportunities. In the current paper we focus on the supply side of the labour market, trying to understand how resources, including human capital, cultural knowledge, social networks, and personality traits, might influence immigrants’ labour market related life course events. We will find that while there is no lack of theoretical ideas in the field of immigrants’ labour market integration, there is indeed a lack of systematic empirical research synthesizing these ideas and trying to provide a genuine empirical test thereof with applicable data and substantive research questions.

4.2.1 Human Capital and Country-Specific Cultural Knowledge

Arriving from abroad, immigrants bring with them educational credentials that are often on a lower level than those in the receiving society or of lower relevance for it. This is due, first, to overall differences in the mean level of educational opportunities in various countries, in particular to disparities between western industrialised countries (host societies) and less-developed countries (sending societies). For example, the lower level of education among immigrants coming from economically depressed areas in Turkey in the 1950s–1970s has been shown to be the prevailing factor in explaining their labour market disadvantages in Germany (Kalter and Granato 2007).

Second, some aspects of human capital, particularly language skills and cultural knowledge,Footnote 1 are country specific, i.e., they are more relevant and productive in some societal contexts than in others (Chiswick 1978; Esser 1999; Friedberg 2000). As a consequence of the act of migration, these aspects are discounted to a certain extent, hence leading to a certain devaluation of human capital. The more dissimilar the structures and cultures of the sending societies are to those of the receiving societies, the more difficult it is for immigrants to adapt to a new society. With the passage of time in the host country, differences between the native-born population and immigrants in terms of human capital should narrow as immigrants learn the host country’s language, gain knowledge about how its labour market functions, and acquire local education and training.

The third point to be mentioned here is that the trend towards assimilation might be slowed down or even disrupted if immigrants are reluctant or unable to invest in the human capital specific to the host country, including the host country’s language (Bonacich 1972; England 1992; Dustmann 2000). Generally, one of the reasons for such reluctance is that immigrants often consider their stay in the host country as temporary, and, reasonably, refuse to make investments that are not certain to pay off (Kalter 2003). Furthermore, employers might also be hesitant to invest in their education or offer on-the-job training, since such investments might be lost if immigrants return to their home countries. For immigrants with pronounced permanent orientation, educational investments by the host country also cannot be taken for granted, as immigrants’ older ages upon migration or the absence of institutional structures in the host country to promote life-long learning for the working age population might represent additional hurdles in this regard.

While the above-mentioned theoretical arguments are by no means new, there are hardly any studies designed to directly test these ideas with appropriate empirical constructs. This is partially so because the official data, e.g., those stemming from the census or labour force survey, are not collected specifically for the analysis of immigrant populations and hence do not contain information on immigrants’ qualifications from abroad, on recognition of those qualifications by the receiving society, or to which degree migrants have participated in retraining courses to gain host-country-specific qualifications. The maximum that can be done with most available data is to establish the highest level of education attained by immigrant respondents and to proxy whether this was acquired in the host country or in the sending country. This procedure often fails to capture nuances of host-country specificity of immigrants’ educational attainment, which might be particularly problematic in countries with a strong signalling value of education for labour market success (e.g., as in Germany).

4.2.2 Social Capital

One of the key notions of life course research is the idea of linked lives or, in other words, the embeddedness of life course events in social relationships (Elder et al. 2003). From the economic perspective as well, social networks are known to be effective sources of information on job vacancies, of reduced transaction costs and of increased efficiency (Granovetter 1974; Waldinger 1996; Burt 1992). Since Granovetter’s seminal work (1973, 1974), a large body of research has addressed the role of social capital in labour market success (Burt 2001; Lin 2001; Lin et al. 2001; Portes 1998). The use of informal search methods is viewed as efficient for both employers and job seekers (Ioannides and Loury 2004; Marsden 2001; Erickson 2001). For employers, referrals by third parties reduce uncertainty related to the value of potential employees’ skills and credentials. For job seekers, the use of social resources provides a means of accessing information on job openings and increases the efficiency of job searches (Montgomery 1992; Flap and Boxman 2001).

The issue of social networks has attracted growing scholarly attention in migration research (see Mouw 2002; Kalter 2006; Martinovic et al. 2009; Drever and Hoffmeister 2008; Haug 2007). Drever and Hoffmeister (2008), for example, show that in Germany almost 50% of all jobs attained by immigrants are found with the help of networks (among the native-born the corresponding figure is 30%). The consequences of these facts are disputed: It is argued that immigrants’ heavy reliance on social networks might lead to lower-quality employment (Kazemipur 2006; Falcon and Melendez 2001; Elliott 2001) or to lower wages (Green et al. 1999). Note that with regard to immigrants, a relevant issue of their networks is not just availability or size, but also quality, particularly with regard to the ethnic composition. If ethnic stratification is in place, homogeneous ethnic networks might be less effective for immigrants’ upward mobility chances and might even lead to a sort of entrapment (Wiley 1970; Portes and Rumbaut 2001; Portes 1998; Portes and Sensenbrenner 1993). On the other hand, and this is particularly relevant for recent newcomers, ethnic community infrastructure and inter-ethnic connections might offer immigrants a shelter in the initial period of their adaptation to the host society and provide security, high solidarity and labour market opportunities within the ethnic economy (Portes 1995; Zhou 1992; Waldinger 1994, 2005; Sanders and Nee 1996; Sanders et al. 2002). According to the segmented assimilation approach, ethnic communities are said to provide an alternative to immigrants’ blocked upward assimilation into the mainstream society, but, what is more important, an alternative to their downward assimilation into the host society’s underclass structures (Portes and Zhou 1993; Zhou 1999). Tight-knit immigrant community structures might cushion immigrants’ economic integration into the middle class, albeit accompanied by lagged acculturation and deliberate preservation of the immigrant community values (see example of Asian Americans, Xie and Goyette 2003).

Although the ideas of social capital analysts are very appealing for the analysis of the labour market inclusion of immigrants, they have rarely been tested in their complexity and under application of appropriate methodological tools. Again, once working with most existing data, one can hardly get any information at all on networks, not to mention information on the ethnic composition of immigrants’ networks.

4.2.3 Immigrant Selectivity on Unobserved Characteristics

The level of educational attainment among immigrants is affected by immigrant selection and self-selection processes, an issue discussed particularly by economists. They argue that people voluntarily emigrating from a particular country are not a random sample drawn from the population, but rather a selected group. Economic immigrants, Chiswick (1978) argues, represent the more ambitious, motivated, risk-taking, and able elements in their source countries. This is so because only persons with such characteristics are willing to take the risky and costly step of migrating. Chiswick’s (1978) explanation for the exceptional success of immigrants arriving in the USA in the 1950s–1960s is, in fact, based on their positive self-selection. Selection clearly varies with respect to the type of migration, economic migrants being more favourably self-selected on the basis of higher intrinsic abilities and economic motivation, and tied movers or refugees less so (Chiswick 2000).

While the above-described patterns of immigrant self-selection are relevant for the comparison of migrants to those who do not migrate, Borjas’ (1987, 1990, 1994) work sheds light on the differences in immigrant self-selection depending on the choice of their destination countries. According to the Roy Model of immigrants’ self-selection (Borjas 1987), immigrants’ selectivity in terms of both observed and unobserved traits depends upon the relative returns on skills in source, respectively destination, countries. Positive selection for skills, i.e., selection from the upper tail of the home country’s income distribution, is expected of immigrants from relatively egalitarian countries, i.e., those in which income is less dispersed, who immigrate to more unequal countries, i.e., the ones with more dispersed income (e.g., the US or Israel for that matter), where highly skilled immigrants can enjoy greater returns on their skills. By contrast, negative selection of immigrants, i.e., those that hearken from the lower levels of income distribution, is expected in the case of immigration from relatively unequal to more equal countries, where the (welfare) state protects the less skilled (e.g., Sweden or Germany for that matter).

The patterns of immigrant self-selection as appealing as they are, are difficult to test directly (Cohen et al. 2008), and many studies have drawn their conclusions ex-post-facto based on the residual differences in economic performance between immigrant and native-born populations after taking observed factors into account (Cohen and Haberfeld 2007). The problem with such analyses is that the residual ethnic or immigrant penalties (Heath and Cheung 2007) might not only be a function of differences in unobservables, but also be driven by other factors, including a difference in institutional characteristics of the host country, e.g., labour market structures, immigration policies, welfare regime institutions etc. (Cohen and Kogan 2007; Kogan 2007). Although hardly any official dataset contains any information on individual unobserved characteristics since these are, per definition, difficult to measure, recent research has made strides in exploring the role of personal traits in labour market success (Heckman et al. 2006; Kunh and Weinberger 2005; Uhlig et al. 2009; Barrik and Mount 1991). Nonetheless, until now hardly any studies have explored the role of personality characteristics in immigrants’ self-selection and their success in the host country’s labour market.

4.3 The German and Israeli Context: On the Role of Institutional Constraints

Migration and immigrants’ integration are interrelated life course events, which are clearly embedded in social institutions and are subject to institutional forces and pressures of both sending and receiving societies (Elder et al. 2003; Mayer 2001). This implies that mechanisms of immigrants’ labour market allocation are mediated and altered by these countries’ institutional constraints. The best way to tease out the role of institutions is to do so from a comparative perspective. In the following we focus on immigrants from the former Soviet Union (FSU) who came to Israel and Germany during the 1990s and early 2000s (for more information on the circumstances of their migration see Gruber and Rüßler 2002; Dietz 2000; Becker 2001; Schoeps et al. 1996, 1999; Cohen and Kogan 2005, 2007). This provides us with an especially interesting and strategic design. On the one hand, immigration of Jewish immigrants from the same source country who were granted a free choice to emigrate to both destination countries enables us to conduct a rigorous comparison of a single ethnic group in two countries whose migration policies, welfare assistance and labour market institutions differ. Jewish immigrants in both countries share a similar ethno-cultural background and demographic characteristics and do not differ significantly with regard to either observed human capital characteristics, or unobservable traits (Cohen and Kogan 2007; Cohen et al. 2008). On the other hand, we can compare Jewish FSU immigrants in Germany, or Jewish quota refugees (JQR) as this group is officially labelled, to Ethnic Germans in Germany. Through this comparison we actually keep the institutional context in the labour market structure and welfare regulations constant, observing two groups that might differ in their individual resources and labour market preferences. Such a comparison might also capture another substantial difference between these two groups of immigrants that is related to the differences in their legal status, including German citizenship acquisition. Whereas JQR have to wait for about 6–8 years to receive German citizenship, immigrants arriving with the status of Ethnic Germans receive it immediately, and their further relatives receive it after 3 years of residence.

While, to our knowledge, no studies have compared JQR and FSU Ethnic Germans in Germany until now, previous research does compare Jewish FSU immigrants across the two countries (Cohen et al. 2008; Cohen and Kogan 2005, 2007; Kogan and Cohen 2007). It singled out the key features of the two countries’ institutional set-ups that are relevant for immigrants’ quick and successful integration into the labour market: the linkage between education and the labour market, particularly the degree of recognition of foreign educational credentials, the flexibility of the labour markets, and the systems of welfare assistance (Cohen and Kogan 2007). All three elements taken together, in Germany we are more likely to observe difficulties on the part of immigrants to enter gainful employment. The link between education and the labour market is one of the strongest in Europe (Müller et al. 1998), and thus detrimental to rapid economic integration of immigrants. At the same time, efforts to recognise foreign educational certificates are minimal, except in the case of Ethnic Germans. Despite recent attempts at flexibilisation, the German labour market remains quite rigid, with relatively high labour costs and a pronounced insider-outsider divide (DiPrete and McManus 1996; Thelen 1991). Finally, welfare assistance in Germany is comparatively generous, possibly constituting a disincentive for highly-educated immigrants to take on low-skilled jobs.

Non-surprisingly, therefore, research on FSU immigrants in the two countries found much higher unemployment risks for FSU Jewish immigrants in Germany than for those in Israel (Cohen and Kogan 2007). However, if able to enter employment, FSU Jewish immigrants in Germany are more likely than those in Israel to enter higher status professional, technical and managerial (PTM) occupations. Cohen and Kogan (2007) conclude that because the German labour market is more rigid and welfare benefits are more generous and long-lasting, FSU Jewish immigrants in Germany, unlike those in Israel, are not forced to take whatever job is available, but can wait for matching employment. Despite these quite plausible explanations, the authors also point to alternative explanations, for example, the different role that ethnic economies might play in the two countries in accommodating immigrant job seekers or their preferences for specific occupations. However, the authors were unable to test any of these ideas due to the lack of adequate data: The official Microcensus data they relied upon does not contain constructs relevant for explaining immigrants’ labour market penalties, discussed above. Furthermore, there are some flaws in identifying Jewish FSU immigrants in the German Microcensus (see Cohen and Kogan 2005, 2007).

In the current paper, we intend to build upon the earlier research of Cohen and Kogan (2007) and try to overcome some of its limits. To this end we present results of analyses based on a primary data collection among the relevant populations in both countries. This study paid specific attention to a proper operationalisation of the above-discussed theoretical mechanisms that could be responsible for explaining immigrants’ labour market penalties.

4.4 Description of the Data and Variables

To answer the questions of whether and why FSU immigrants vary in their chances to become successfully integrated into the receiving societies’ labour markets we use data from a comparative survey in Germany and Israel. In both countries we interviewed immigrants aged 25–54, who arrived from the FSU at age 18 or older in the years 1994–2005. In Germany the survey was conducted by telephone between May and June 2007 solely among immigrants and the data for the native-born German reference population was supplemented by the national representative sample from the German Socio-Economic Panel (GSOEP), wave 2007 (see Liebau 2010).Footnote 2 In Israel data was also collected per telephone, both for FSU Jews (interviews were also conducted in Russian) and native Israelis in August 2007 and April 2008. The total number of cases collected in Germany is 655 FSU Jewish immigrants and 891 FSU ethnic Germans (Germany); and in Israel, 827 FSU immigrants, and 628 native Israelis (Israel).

For the sake of comparability, we use the same dependent variables as in the earlier research (Cohen and Kogan 2007): unemployment risk and the propensity of gaining employment in professional, technical and managerial (PTM) jobs. The former dependent variable differentiates between unemployed individuals and those in employment, thus focusing on the active labour force. The latter dependent variable refers to the propensity of gaining employment in professional, technical and managerial jobs, which are the jobs pertaining to groups 1–3 (encompassing legislators, senior officials and managers, professionals, technicians and associate professionals) on the International Standard Classification of occupations (ISCO-88). In addition, once analysing immigrant populations in both countries we also explore immigrants’ chances to flock to ethnic economies. Immigrants are considered to work in the ethnic economy if the proportion of co-ethnic colleagues or employers at the respondent’s working place is more than half. All dependent variables pertain to employment status at the moment of observation, i.e., in summer 2007. Due to the binomial character of all three dependent variables we use binomial logistic regressions as our basic tool of analysis.

As independent variables we also replicate those used in the study by Cohen and Kogan (2007), i.e., gender, age, family status, and education. Most importantly, however, we are able to look at further indicators capturing the constructs of the major theoretical arguments. We control for host-country-specific human capital, capturing all relevant aspects, such as recognition of foreign educational credentials and participation in schooling and/or training in Germany or Israel. Host-country-specific cultural knowledge was measured by means of language proficiency, both subjectively assessed and the fact of speaking German/Hebrew at home. Social networks are captured in several ways. First, we explore whether immigrants had friends or relatives in Germany, respectively Israel, before immigration. Second, we assess the composition of respondents’ network of friends (Russian/German/Israeli) at the moment of survey. All these immigrant-specific variables are coded in such a way that a value of zero (the reference category) reflects immigrants’ resemblance to (or assimilation with) the natives. The value for natives was set to zero anyway, so that in a sense we estimate an immigrant-specific interaction effect. For example, immigrants who do not speak the language of the host country at home are assigned a value of 1 in the variable ‘Not speaking German/Hebrew at home’. Immigrants who speak German/Hebrew at home are assigned 0, and 0 is also assigned to all natives. Thus, the effects of all immigrant-relevant variables discussed above pertain to the gap between immigrants with varying characteristics on these variables, whereas the main effects for the immigrant groups pertain to the average differences between these groups and the natives.

Further, we control for personality traits measured by the so-called ‘Big Five’ (Dehne and Schupp 2007; Lang and Lüdtke 2005), including items pertaining to openness, conscientiousness, extraversion, agreeableness, and neuroticism. By including these measures we try to gain insight into the importance of psychological factors that are potentially relevant for immigrants’ labour market integration but are usually ignored in the migration research. The variables capturing character traits are controlled in the study, but we do not estimate effects of personality traits separately for immigrants and natives, as this would overstretch the current paper. Hence the effects of personality traits are controlled for in the models but not reported separately in the tables below.

In models comparing FSU immigrants across countries, or JQRs and Ethnic Germans in Germany (as in the analyses presented in Table 4.3), we also control for years since migration and for the latter models also for citizenship status, contrasting immigrants possessing German citizenship with those who are not yet naturalized.

The analyses include only cases with non-missing information on all relevant variables. This results in 530 valid cases for JQRs, 730 valid cases for Ethnic Germans, 6,765 valid cases for the native Germans, 670 valid cases for FSU immigrants in Israel, and 566 valid cases for native Israelis. For the description and definitions of all variables used in the analyses see Table 4.A1 in the appendix.

4.5 Results of the Multivariate Analyses

4.5.1 Extending the Set of Individual Resources as Predictors of Gainful Employment: Jewish FSU Immigrants Compared to Natives in Germany and Israel

We start our multivariate analyses by replicating the model used in the earlier analyses with the standard set of independent variables (like gender, age, age squared, marital status, and education). To this end we only compare Jewish FSU immigrants to natives in both countries, running separate regression models, neglecting Ethnic Germans for the time being. The results are presented in Model 1 in Table 4.1; they largely correspond to the findings found in Cohen and Kogan (2007).Footnote 3 The odds-ratio of 9.07 signals that Jewish FSU immigrants in Germany face a much higher risk of unemployment (relative to the natives) than their counterparts in Israel, for whom we find hardly any disadvantage with regard to employment (odds-ratio: 1.17). When looking at the relative chances of the immigrants to enter PTM employment, we find huge disadvantages of comparable size (0.25, resp. 0.29) in both countries. Thus a central finding from the earlier studies is confirmed by replicating the analyses with our new dataset.

Table 4.1 Relative labour market success (compared to natives) of Jewish FSU immigrants in Israel and Germany; odds ratios (and standard errors) from country-specific logistic regressions
Full size table

In Model 2, we include a larger set of independent variables identified as relevant for explaining immigrants’ labour market chances. Comparing the odds-ratios to those in Model 1 shows that taking into account differences between immigrants and natives in relevant country-specific human capital, cultural resources, social networks and personality characteristics; we are able to account for a large part of the immigrants’ penalties. The formerly ninefold penalty of FSU Jewish immigrants in Germany (compared to the charter population) is now reduced to a 2.3-fold penalty, albeit remaining statistically significant. In Israel the above-mentioned constructs do not seem to play any role in explaining immigrants’ labour market chances, which is not particularly surprising since we observe hardly any employment penalties for the FSU Jewish immigrants there. Model 2 is particularly successful in explaining penalties with regard to the job status of employed immigrants. Clearly, the variables added in Model 2 are crucial to understanding the relatively low occupational success of Jewish FSU immigrants in both countries. Once we observe immigrants with characteristics comparable to those of the native-born, the penalties with regard to PTM employment that are visible in Model 1 disappear. Now, in Model 2 we observe even somewhat higher (albeit insignificantly) relative chances among immigrants to land PTM jobs (1.79, resp. 1.32) in both countries.

Our next table, Table 4.2, shows how the central constructs of interest, i.e., human capital, cultural capital, and social capital, operate in each of the two countries with regard to both dependent variables. In general and consistent with the findings from Table 4.1, it is evident that these predictors make a more pronounced contribution to explaining immigrants’ penalties in Germany than in Israel. If immigrants’ FSU education is not recognised in Germany, they are less likely to face unemployment (0.60), but also have much lower chances of entering PTM employment (0.34). Whereas the latter effect is quite understandable – in the end it is difficult to enter high-status employment in a host society if one’s education is not officially recognised – the first finding appears less intuitive. It might indicate a voluntary aspect of unemployment among those whose FSU education is recognised. The finding that they are more likely to be unemployed might be related to the fact that these people wait in order to enter matching employment, a phenomenon already mentioned in Cohen and Kogan (2007). In Israel official recognition of education does not entail any obvious advantage to those who gained it (1.19, resp. 0.96).

Table 4.2 Effect of individual resources in Israel and Germany; odds ratios and standard errors in parentheses
Full size table

Host-country-specific education and training seems no less important for securing employment, particularly higher-status employment. In Germany immigrants who did not invest in host-country education are penalised by higher chances of unemployment (1.83) and somewhat lower chances of entering PTM employment (0.58; only significant at the 10 percent-level). A lack of Israeli education and training hardly affects the unemployment risk of FSU Jewish immigrants (1.32), but it dramatically lowers their chances for PTM employment by the factor of five (0.20).

Lack of proficiency in the language of the host country matters, irrespectively of its measurement (subjective or objective), but it matters significantly only in Germany. FSU Jewish immigrants with a poorer command of spoken and written German are more likely to face unemployment (1.66) and have substantially lower chances of entering PTM employment (0.29). Similarly, not speaking German at home increases the unemployment risk enormously (4.13) and also reduces the odds of entering PTM employment (0.38) in Germany. Although these results are in line with our theoretical predictions, it is nonetheless puzzling why we do not find similar effects for the knowledge of Hebrew in Israel. Apparently the fact that knowledge of Hebrew does not appear to determine chances of FSU Jews either for any employment or for higher-status employment is related to the opportunities open for them in Russian-speaking ethnic economies, which is quite pronounced in Israel (Mesch 2002).

Our results with regard to the role of the social capital are somewhat disappointing. Apart from a single effect of not knowing someone in Germany prior to migration, which seems to increase the unemployment risk of FSU Jews there (odds ratios of 1.57), no other social-capital variables turned out to have a strong effect or to be significant. Specifically in Israel the social network of FSU Jews does not seem to play any particular role in easing their employment chances. It should be noted, however, that the two variables included in these analyses are quite indirect measurements for the use of social contacts as a means of securing access to any employment. Simply knowing someone in the host country prior to migration does not mean that such contacts will be helpful in providing information about job opportunities or even more – in serving as references for suitable jobs. Similarly, a simple account of the ethnic composition of the friendship network does not say anything about whether this network will be helpful in providing access to gainful employment. Moreover, the issue of causality remains unresolved here, as it might equally be that the employment status affects the composition of one’s friends’ network and not the other way around. That is why in the future research, focusing solely on the FSU immigrants in Germany, we will rely on more direct measures of social networks that might be more telling for the labour market success and apply methods of analyses more appropriate for detecting causal relationships (e.g., event-history modelling).

4.5.2 Pursuing the Effects of Institutional Restrictions: Comparing Jews in Germany to Jews in Israel and to Ethnic Germans in Germany

Thus far our analyses have concentrated on the comparison of FSU Jewish immigrants to the native-born across countries. After adequately capturing differences in the individual resources between immigrants and natives and accounting for usually ‘unobservable’ characteristics by including information on central personality traits, we are more certain to relate at least part of the existing ethnic penalties to the institutional structures of labour market regulation and to the welfare assistance in place in the immigrant-receiving societies. We were indeed able to show that after controlling for individual resources carefully, ethnic penalties exist solely in Germany and only with respect to immigrants’ employment access. There, FSU immigrants with relevant human capital, cultural capital, social capital, and personality characteristics are still less likely than the native-born to enter employment.

Additional insights can be expected from a direct comparison of FSU Jews across the two countries. This means that we are no longer comparing differences (between countries) in differences (of immigrants to natives) anymore, but are looking directly at the differences in the employment chances of FSU Jews in Germany and those in Israel. In a sense, this follows the idea of a natural experiment: indeed all those Jewish immigrants (and their non-Jewish family members) who ended up in Germany could as easily have entered Israel and for most in Israel the opposite is true. In such a comparison we ask how FSU Jewish quota refugees who headed to Germany would fare in the Israeli labour market, had they chosen Israel instead of Germany as their country of destination.Footnote 4

Another meaningful comparison is the comparison of Jewish quota refugees in Germany to another FSU immigrant group found there: Ethnic Germans. Both immigrant populations face the same labour markets. They differ however in their privileges, such as entitlement to citizenship and recognition of education certificates, as well as in their ethnicity and pre-migration socio-demographic characteristics, which might determine their preferences and job aspirations. If these two groups still differ in the patterns of their labour market incorporation once we have adequately controlled for the apparent socio-demographic variation between the two groups, this points to the importance of the legal institutional context in Germany.

Table 4.3 presents the results of these additional comparisons. As in the earlier analyses, we look at unemployment risk and propensity for PTM employment as dependent variables. Further, we extend our analyses to an additional dependent variable: employment in an apparent ethnic niche. The first column, which shows the direct comparison of the labour-market chances of FSU Jewish migrants across countries, delivers a well-expected pattern: compared to FSU Jews in Israel, Jewish quota refugees with similar characteristics (see the extensive set of controlled variables) in Germany face significantly higher unemployment risks (3.42), and have somewhat higher chances to enter PTM-employment (1.43, the effect is significant at the 10 percent-level), but dramatically lower chances to end up in the ethnic economy. As expected, at least some of the FSU immigrants in Israel are absorbed by opportunities available in ethnically organised businesses, whereas in Germany the ethnic economy apparently creates much fewer job possibilities for FSU immigrants. This is by no means surprising, taking into account an overall higher proportion of FSU immigrants in Israel, particularly in relation to the country’s total population.

Table 4.3 Jewish FSU immigrants in Germany compared to Jewish FSU immigrants in Israel and Ethnic Germans in Germany; odds ratios and standard errors in parentheses
Full size table

The second column in Table 4.3 allows us to compare Jewish quota refugees to Ethnic Germans within Germany. Results clearly show that Jewish quota refugees, other things being equal, are more likely to be unemployed, but at the same time are significantly more likely to enter PTM jobs if successfully employed. These results point against the assumption of possible discrimination against Jewish immigrants in Germany – why should JQR be discriminated at job entry but preferred when it comes to higher-status employment – in favour of another explanation. As already mentioned elsewhere (Cohen and Kogan 2007), we might be dealing with a voluntary aspect of unemployment among FSU Jewish immigrants in terms of their ‘waiting for matching jobs’. Finally, results show that Ethnic Germans, who are a much larger group of FSU immigrants in Germany, do not use opportunities of the Russian ethnic economy any more than do JQR to enter gainful employment.

4.5.3 Accounting for the Residual Effects: Preferences or Aspirations?

So far we were able to confirm previous findings of a higher risk of unemployment among FSU Jewish immigrants in Germany than among native-born Germans, Ethnic German immigrants, or FSU Jews in Israel. Controlling for an extensive set of independent variables pertaining to individual human, cultural and social capital as well as to personality characteristics, we are able to account for a large part of the employment penalty, but a residual effect still remains. How can we explain the fact that Jewish quota refugees in Germany are more likely than Ethnic Germans to stay unemployed, apparently awaiting better employment chances and eventually getting them?

In the theoretical sections we mentioned that the job search behaviour of the native-born might differ from that of immigrants. Being rational actors they are expected to enter employment if it promises them labour-market returns higher than those of their status quo upon remaining unemployed. If highly skilled immigrants face difficulties in entering adequate employment in a country with relatively high levels of welfare assistance (which is higher in Germany than in Israel, see Cohen and Kogan 2007), they might rationally prefer to stay unemployed, as long as they are sure in the end to enter high-status employment matching their qualifications. In other words, such immigrants might maintain their high reservation wages, which prevent them from entering employment at any price. If the labour market does not support such practices of ‘waiting’ but instead pushes immigrants, even highly skilled ones, to enter any employment at all, then immigrants will not stick to their high reservation wages but will enter employment more quickly. The latter situation is more likely to occur in Israel than in Germany.

In Table 4.4 we report reservation wages for unemployed immigrants and the native-born in the two countries. The figures for the native-born and FSU immigrants in Germany are taken from different datasets and thus are not strictly comparable. Furthermore, the results should be approached with some caution, as the sample sizes are quite small. Even if the results suffer from these and other flaws, they point in the expected direction. Compared to natives, FSU immigrants in Germany, particularly FSU Jewish quota refugees, declare higher wages needed for them to enter employment. This is particularly evident in the contrast between FSU Ethnic Germans and FSU Jews, which does not suffer from limitations of incomparability and is quite telling: reservation wages of FSU Jewish immigrants are substantially higher than those of Ethnic Germans, and the gap between the two remains pronounced, even if we look only at the highly educated.Footnote 5 The picture is opposite in Israel. Here, in accordance with our expectation, FSU immigrants claim to have somewhat lower reservation wages than the natives. Our results thus deliver one answer to the original question of why FSU Jews in Germany might face higher unemployment. Apparently they rationally decide to wait for better jobs, which are not quickly obtained in a rigid German labour market.

Table 4.4 Reservation wage for unemployed FSU immigrants and natives
Full size table

It still remains unclear why FSU Jewish immigrants allow themselves to wait for suitable employment at times when Ethnic Germans settle for jobs of much lower occupational status while claiming much lower reservation wages. Welfare state differences cannot be an explanation here, because welfare assistance extended to Ethnic Germans is not much different from the opportunities that Jewish quota refugees can make use of.

Apparently we should search for alternative explanations as well, one of those mentioned earlier being job aspirations. If we turn to the life course of FSU immigrants back in their home countries and look at the list of the most frequent occupations Ethnic Germans and Jewish quota refugees used to pursue in the former Soviet Union, the difference is striking (Table 4.5). JQR used to work in high-status highly specialised occupations for which special licences (as in the case of medical doctors) or German citizenship (as in the case of teachers, who in Germany hold a special status of civil servants) are necessary. Having spent many years in acquiring the qualifications necessary for these professions, JQR might be reluctant to give them up easily and instead work in lower-status occupations, which in Germany they might be able to enter with fewer obstacles. JQR are apparently more likely to engage in further education and training as well as German language courses to improve their chances for higher status employment (see also Table 4.A1 for this evidence). Unlike JQR, Ethnic Germans tended to work in the FSU more in blue-collar or lower-level white-collar occupations, for which retraining is not that essential, high language proficiency is less of a precondition and status loss is not that pronounced.

Table 4.5 Five most frequent occupations among the German FSU immigrants in their home countries
Full size table

4.6 Summary and Discussion

The major aim of this paper was to explore the importance of individual resources and of structural constraints for an important life course event among immigrants, their successful labour-market integration in a new country. We focused on human capital, above all on host-country-specific education, cultural capital in terms of proficiency in the language of the host country, social capital in terms of composition of immigrants’ friendship networks, and personality traits. We also pointed to crucial differences in the receiving contexts of Germany and of Israel. Although the importance of the above-mentioned dimensions is theoretically indisputable, to date surprisingly little empirical work has been able to analyse the whole range of individual resources in one empirical model. Very few surveys contain information on a large set of variables directly related to the theoretically derived constructs, and almost no study explores this information in a strictly comparative design. The current study, contributing to theory-driven empirical research, with a substantive interest in explaining differences in labour-market attainment among FSU immigrants in Germany and Israel, is an obvious exception.

In accordance with already existing research, the current study confirms that FSU immigrants and native Israelis do not differ substantially in their employment propensity. Furthermore, there hardly any differences between the two groups in their propensity for PTM employment once completion of local education and training is achieved. The situation differs dramatically in Germany. Here, FSU Jewish immigrants have a higher propensity towards unemployment, even after controlling for their human, cultural, and social resources and character traits, although the disadvantage declines strongly when these individual characteristics are taken into account. After including a detailed set of individual resources we are even able to fully explain immigrants’ disadvantage with regard to the status of the jobs they attain. In fact, unlike in prior analyses based on a limited set of predictors (Cohen and Kogan 2007), in our current analyses for Germany FSU immigrants exhibit a similar propensity for PTM employment, given completion of host-country-specific education and training and a high level of language proficiency.

While we are able to account for large parts of the immigrants’ penalties with respect to employment chances, we are still unable to explain why FSU Jewish immigrants face a remaining twofold risk of unemployment in Germany, even when controlling for the extended set of independent variables. Searching for exhaustive explanations, we confirmed that the existence of an extensive ethnic economy in Israel contributes to an easier labour-market entry of Russian Jewish immigrants there. Apparently, the fact that the Russian ethnic economy is not that pronounced in Germany (see Table 4.A1 for such evidence), not the least due to the heterogeneity of various FSU immigrants settling there, keeps many FSU immigrants out of employment, particularly those who are less oriented towards the receiving society.

Another part of the explanation for existing cross-national differences in employment chances of the FSU Jews seems to be the existing welfare support systems in the two countries. Whereas the German welfare system is erected to protect individuals from occupational downgrading, and despite its very recent amendments still allows them enough time to find employment matching their educational qualifications, the minimalist Israeli approach encourages the unemployed to take up any employment, often at the price of a severe occupational mismatch. These assumptions were indirectly confirmed in our study, as we were able to show that the lower unemployment risk among FSU immigrants in Israel (compared to Germany) is partially attributed to the lower reservation wages among FSU immigrants. Instead FSU Jewish immigrants in Germany have higher hopes of finding employment suitable to their occupational qualifications, definitely more so than Ethnic Germans from the FSU.

Education, particularly education obtained in the host countries, appears to be one of the strongest predictors of immigrants’ labour market success. Our evidence confirms this for both the German and the Israeli case. However, we started this paper with the claim that accounting for human capital alone by far is not sufficient to explain immigrants’ labour market penalties in modern immigrant societies. Cultural knowledge and social resources are necessary in order to enter gainful employment. Our findings seem to strongly support this view, particularly in the German case. Although in Germany educational credentials remain the most important predictor for job entry and occupational mobility, if such qualification are not attained or at least not recognised in Germany, migrants are bound to face serious difficulties with regard to employment. Language problems apparently also contribute to immigrants’ weak labour market performance, but more so in Germany than in Israel. Social resources matter then, albeit to a lesser degree than expected.

Have all questions been answered now? Certainly not, as we at least also have to take into account the demand side of the labour market exploring whether German employers do not favour Ethnic Germans (Aussiedler) and discriminate against JQR, the assumption we have refuted for the time being. Future research should certainly address the question of whether German employers are at all able to differentiate between the two groups and do not just consider all immigrants from the FSU as “Russians”.