Mirna SAFI Observatoire Sociologique du Changement (OSC) Sciences-Po/Cnrs and Laboratoire de Sociologie Quantitative (LSQ) Crest/Insee Author’s address Observatoire Sociologique du Changement (OSC) Sciences-Po/Cnrs 27 Rue Saint-Guillaume 75337 Paris Cedex 07 France Tel. +33(0)1.45.49.54.78 Fax: +33(0)1.45.49.54.86
[email protected]
Immigrants’ life satisfaction in Europe: between assimilation and discrimination Abstract Although several recent studies have provided insights into European attitudes towards immigration, very few have explored the attitudes of immigrants themselves and their perception of their own lives in the host country. Using data from the three rounds of the European Social Survey, this article investigates disparities in life satisfaction measures between immigrant first and second generations on the one hand, and natives on the other hand in thirteen European countries. Two major theoretical hypotheses explaining the lower level of immigrants’ subjective well-being are tested: assimilation and ethnic penalties. The main finding is that migration and the assimilation process do not account for immigrants’ relative dissatisfaction. When ethnic groups are compared, the discrimination some of them perceive in the host society seems to be a more consistent explanation for their lower life satisfaction level. Unlike other studies, the effect of discrimination is measured here with an attempt to correct for the endogeneity bias it may lead to by using simultaneous regressions with instrumental variables.
Immigrants’ life satisfaction in Europe: between assimilation and discrimination Based on the exploitation of European comparative data, this article makes the use of life satisfaction measures in order to provide a sociological diagnostic of the immigrants’ assimilation process in some European countries. First, some theoretical hypotheses are presented in an attempt to explain the subjective differences in well-being among ethnic groups. A measure of these differences among immigrants, immigrants’ descendants and the host society for thirteen European countries is then presented. Finally, this work tries to relate the low level of life satisfaction of immigrant first and second generations to their experience of discrimination.
Theoretical perspectives While measures of happiness and contentment with one’s life have been compared at length between countries and social classes, very few studies focus on their disparities among ethnic groups and even fewer are the ones that link these disparities to the socio-psychological assimilation of immigrants. A review of the literature suggests two possible schemes of explanation for the differences in reported well-being between immigrants and natives. The first one links immigrants’ life satisfaction to the acculturation process and the complex psychology of migration. The second one puts the stress on the inferior living conditions of immigration and the discrimination they may perceive in host societies. From the psychological point of view, migration, establishment in a new country and assimilation of immigrants go together with sorrow, melancholy and despair (Park 1928 ; Handlin 1951 ; Handlin 1966). Concepts like acculturative stress have been used by some studies to designate these immigration “psychopathology” (Berry et al. 1987 ; Berry 2001). Immigrants’ psychological adaptation is thus regarded as a long term outcome of intercultural interaction caused by migration. Nevertheless, an underlying hypothesis of assimilation is that this pain is supposed to diminish over time and generations. The straight line conception of assimilation attaches indeed a great importance to duration both from the individual and generational point of view. The assimilation paradigm supposes that immigrants become more and more similar to natives as years go by, and that assimilation progresses with the thread of immigrant descendant generations (Abramson 1994 ; Alba 1995 ; Alba and Nee 2003). However, the importance attributed to acculturation in explaining immigrants’ wellbeing in the host society, or even more generally their assimilation process, has been challenged on theoretical and empirical grounds. Many studies put indeed the stress on the role of immigrants’ objective living conditions in the explanation for their lower subjective well-being (Cummins 2000). They rather emphasize structural assimilation as defined in Milton Gordon’s classical work (Gordon 1964). Gordon’s main argument is that acculturation is likely to take place first within the assimilation process and that a stage of “acculturation only” can continue without any of the other types of assimilation, such as identification or structural assimilation, occurring. Gordon’s scheme thus suggests that migrants can be acculturated but still experience durable “inferiorization” mechanisms and may be therefore dissatisfied from their objective life conditions in the host country. Many research works studied this structural inferior position of immigrants in the social stratification, speaking sometimes of ethnic penalties (Castles and Kosack 1973 ; Carmichael and Woods 2000 ; Heath and Yu 2005), and some other times of downward or segmented assimilation (Portes 1
and Zhou 1993 ; Rumbaut 1997 ; Zhou 1997 ; Safi 2008). An important point is that, if their life chances are perceived and anticipated as sharply and unfairly inferior by the immigrant populations, dissatisfaction can last, if not be reinforced, over time and generations. A key variable plays indeed an important role in this framework; the experience and perception of discrimination. In a well-known article on life satisfaction differences between Blacks and Whites in the United-State and in their query of explanations for continuing disadvantage of Blacks compared to Whites, Thomas and Hugues concluded that “discrimination is a plausible though untested explanation for the differences we have documented in this paper in psychological well-being and quality of life” (Thomas and Hughes 1986, p. 840). According to the authors, this conclusion still held a decade afterwards and had been reinforced by empirical evidence (Hughes and Thomas 1998). Indeed, epidemiologists, social psychologists and sociologists have investigated the effects of discrimination on individual well-being. A direct relation had been demonstrated between perceived discrimination and mental health, social stress or even depression, especially for young ethnic and racial minorities (Vega and Rumbaut 1991 ; Neto 1995 ; Finch, Kolody, and Vega 2000 ; Neto 2001 ; Taylor and Turner 2002 ; Sellers et al. 2003). Based on a sample of 5000 immigrant children in Southern California and South Florida, an investigation conducted by Rumbaut found that perceived discrimination increases depressive symptoms and that expected discrimination is significantly associated with a decrease in self-esteem (Rumbaut 1994). Some studies in social psychiatry and epidemiology suggest that the effect of discrimination on psychological well-being operates indirectly through diminishing self-efficacy and increasing stress. Ethnic differences in depression symptoms are accounted for by the role discrimination plays in producing feelings of helplessness and despair (Hughes and Demo 1989). As Hugues and Thomas suggested, the negative effects of perceived discrimination should be interpreted in social psychological terms rather than in psychiatric ones. Ethnic and racial groups do not register a higher level of mental disorder than Whites, once the social characteristics are controlled for. Reported subjective well-being should be interpreted as the respondents’ evaluation of the quality of their life experience. Perceived discrimination affects the anticipation or the perception members of minority groups may have of their life chances and therefore depress their psychological well-being. Very few studies were conducted on the psychological well-being of immigrants and their descendants in Europe. Some isolated research works focused on the situation in a specific country or for a specific ethnic minority (Neto 1995 ; Werkuyten and Nekuee 1999 ; Neto 2001 ; Baltatescu 2005). The purpose of this article is to compare the situations of first and second immigrant generations in some European countries through their own global evaluation of their lives. Immigrants’ well-being is compared to natives’ one, when the objective differences are taken into account. Then, if differences are still significant, what interpretations can be put forward? What do immigrants’ well-being measures tell us about the assimilation process? And what do they tell us about the discrimination these populations experience in the host country?
Data The data used combine information available from the three rounds of the European Social Survey (ESS); the first was conducted in 2002, the second in 2004 and the third in 2006. ESS is a cross-sectional multistage survey carried out on probability samples of individuals aged 15 and older and comparing more than 20 European countries1. Only countries for which data from the three rounds are available and with sufficient observations for both immigrant 1
The number of countries varies along with the ESS rounds.
2
individual and their descendents are included in the analysis. It is the case of the countries that have been experiencing a long standing immigration: Austria (AS), Belgium (BE), France (FR), Denmark (DK), Germany (DE), the Netherlands (NL), Norway (NO), the UnitedKingdom (GB), Sweden (SE) and Switzerland (CH). Though immigration to Portugal (PT), Spain (ES) and Ireland (IE) is more recent, it has been occurring fast enough to lead to large sizes of immigrant groups. They are therefore introduced in the analysis. Information about the respondent’s country of birth and the one of his/her parents was used in order to distinguish among three different sub-populations with regard to the link to migration. The first generation is the one that migrated (i.e. the foreign born population excluding expatriates). The second generation is composed of native individuals whose parents were born abroad, while natives who do not have any foreign born ancestor form the third generation. Within the second generation, a distinction is made between those whose both parents are foreign born (G2) and those who have only one foreign born ascendant (G2.5). In order to reduce the magnitude of the age effect in the analysis below, only individuals aged between 18 and 65 are included in the sample. Figure 1 depicts the principles of the categorisation of the populations into “generations stemming from migration” while table 1 gives some general figures on their relative importance in the sample.
Figure 1 Categorisation of ESS population Born in country
YES
NO
Both parents born in country
Generation 3
Both parents born in country Both parents born abroad
One parent born abroad
Generation 2.5
Generation 2
Yes
Generation 3 (native expats)
3
No
Generation 1
Table 1 Figures of immigration in the ESS sample
Source ESS round1, 2 and3
Findings In each of ESS rounds a question about individual life satisfaction was systematically asked. •
“All things considered, how satisfied are you with your life as a whole nowadays?
Answers to these questions were given in a scale from 0 to 10 (0 meaning extremely dissatisfied and 10 extremely). Table 2 Life satisfaction of immigrant generations Life satisfaction Mean
Std. Err.
G1
7.01
0.03
G2
6.89
0.06
G2.5
7.29
0.04
G3
7.28
0.01
All
7.25
0.01
Source ESS round 1, 2 and 3
Immigrants and second generation individuals whose both parents are immigrants (G2) report lower levels of life satisfaction. Individuals belonging to G2 seem to be even less satisfied with life as a whole than the first generation immigrants. However, individuals whose only one parent is immigrant report slightly higher levels of both happiness and life satisfaction than natives who do not have any immigrant ancestor. The next section tries to provide possible explanations of these disparities. Regression analyses are estimated in order to compare natives and immigrants’ well-being taking into account three sets of co-variables: sociodemographic factors (gender, age, family status and years of education2), socio-economic factors (income, employment status and occupation) and health factors (subjective response about health status). Different types of ethnic variables are tested: citizenship and place of birth, immigrant generations, and finally the ethnic group. 2
The variable of education level was not used because it has been reported differently in the United-Kingdom. The results estimated when controlling for the highest level of education without data for the United-Kingdom are very close to the ones presented here.
4
Table 3 Life satisfaction determinants: the effects of the control variables
Source ESS round1, 2 and 3
5
Effects of the control variables As expected, important differences are observed across countries: the country parameters are always highly significant (table 3). With France picked as a reference, and with the exception of Portugal, living in any of the other countries selected in this study is associated, other things being equal, with a higher level of life satisfaction. The highest parameters are those for Denmark, Sweden and Switzerland, as reported by the international studies on this subject. ESS rounds do not seem to affect reported well-being. This is probably due to the fact that the duration of the survey is short (2002 to 2007). The effects of the three sets of variables developed above seem to be very stable and conform to those reported in the literature on the determinant of life satisfaction (Frey and Stutzer 2002). Women are more satisfied with their lives than men. The age effect is non linear; the youngest and the eldest reporting higher level of well-being. Life satisfaction increases with annual income. The rise is however more important in the modest and medium brackets than in the high ones. People who do not report their income value seem to be in a rather favourable situation. As for occupation, when compared to unemployment, all occupations are associated with higher levels of life satisfaction and this is all the more true for the most advantaged ones. Moreover, the unemployment’s negative effect seems to be lasting; individuals who have experienced unemployment during the last three months before the survey have a lower probability of being satisfied with their lives. Finally, the health effect is very significant; it becomes very important for those who asserted to have a very bad health situation. The introduction of the subjective perception of health can be criticised: one may think that there are some individual unobserved characteristics (personality ones) that affect both the probability of reporting low satisfaction and a bad health situation. Nevertheless, the introduction of a measurement of health, even if biased by this subjective aspect, remains important in the study of life satisfaction determinants: for instance it corrects for the strong negative effect that would appear for the eldest categories if health was not controlled for. As far as the results related to the immigration issue are concerned, it should be noted that the omission of the health variable does not change the final results of this research. Testing for immigrants’ assimilation According to the assimilation framework, differences between natives and immigrants in life satisfaction levels should diminish with the latters’ length of stay and over the successive immigrant generations. Model 1 in table 4 begins by testing the migration consequences; as expected its effect on individual well-being is significantly negative. Model 2 shows that this negative impact tends to diminish when immigrants have access to citizenship through naturalisation3. If migration does have a negative effect, does this effect diminish with the assimilation progress? Models 3 and 4 provide a test for two important components of assimilation: immigrants’ length of stay and immigrants’ generations.
3
In the following analysis the naturalisation variable won’t be introduced because of colinearity problems. When the aim is to compare first and second generations, clear problems of colinearity with the citizenship variable occurs; except for Germany and Switzerland, second generations are almost always citizens in the European countries studied here, since they are native born. Colinearity becomes also problematic when trying to measure the length of stay variable; immigrant citizens are most probably those who have spent a long duration in host country. Nevertheless, it should be noted that the models estimated with the naturalisation variable give very similar results.
6
Table 4 Life satisfaction determinants: the effect of the assimilation variables
Source ESS, round 1, 2 and .3. OLS regression results, the control variables are included in the analysis (see table 7 in the appendix)
a)Years since migration As accounted for in many studies, the first year of migration seems to be a “happy” one (model 3, table 4). The level of immigrants’ well-being decreases afterwards and remains low for a long time. The Fisher test for equality of coefficients asserts that estimates for 1-5 years, 5-10 years and 11-20 years are not significantly different. Immigrants’ well-being improvement over time cannot be thus discerned from the data. One has to wait for twentyyear migration duration to measure a significant increase in immigrants’ life satisfaction level. Furthermore, even after more than twenty years spent in the host country, immigrants still register significantly lower life satisfaction than natives. b) Inter-generational assimilation The inter-generational assimilation hypothesis does not appear to stand either. When the second generation is defined by having at least one immigrant parent, consistent improvement in life satisfaction seems to take place over generations (table 4, model 4). However, a more precise model (table 4, model 5) which compares G2.5 and G2 reveals that increase in life satisfaction occurs only for those who have one native parent (G2.5). Even after controlling for the demographic and socioeconomic variables, the second generation individuals whose both parents are immigrants still have lower level of life satisfaction than natives with non immigrant parents. Moreover, the Fisher test for equality of coefficients indicates no significant difference between the coefficients of G2 and G1. While born in the host country, descendants of migrants do not seem to be “happier” than their parents, other variables controlled for. The migration hypothesis does not hold in explaining their relative despair since they are native born and did not migrate ever. No evidence is detected for the intergenerational assimilation hypothesis either. These findings suggest that, as far as individual well-being is concerned, the “psychology of immigration” framework does not provide satisfactory explanations for disparities between immigrants and their descendants on the one hand and natives on the other hand. Moreover, these findings also highlight the importance of the distinction between G2 and G2.5 in immigration studies. Descending from a mixed couple
7
has been shown to facilitate the integration of second generations, namely by increasing their social capital and thus their labour market outcomes (Rooth and Ekberg 2003 ; Ramakrishnan 2004). In the present article, the results indicate that if one of their parents is native born, second generation immigrants also have a better level of life satisfaction. It is thus not the course of inter-generational assimilation that affects the second generations’ “psychology”, but rather something that seems to be related to the ethnic group they belong to. Ethnic minorities and subjective well-being If neither the length of stay nor inter-generational assimilation affects immigrants’ well-being, the ethnic group membership and its relation to subjective perceptions may make the difference in explaining life satisfaction disparities between immigrants and natives. All immigrant groups do not experience the same living conditions in host countries and some of them may be characterised by durable ethnic penalties that entail their lower life satisfaction. In order to test this “ethnic penalties” hypothesis, first and second generation immigrants are distinguished by their country of origin. Unfortunately, the parents’ country of origin is not given in details in the first ESS round. Only information about the geographic area is available with the regrettable lack of distinction between eastern and western Europe. However, the detailed country of origin variable is available for the second and third ESS waves. Therefore, distinct analyses will be conducted first on the three rounds and second on the last two ones only. When all available rounds are used, we distinguish between six ethnic groups coming from Africa, Asia, South-America, Europe, other geographical zones and “non ethnic” natives. When the study is concerned only with the two last ESS rounds, one can distinguish between Eastern and Western Europe. In the latter case, special attention has been also drawn to individual with Turkish origins. We classify first and second generations in each of these ethnic groups according to the information about their country of origin and/or their parents’ one. Second generation immigrants descending from mixed couples are classified according to the origin of their immigrant parent. When they have two immigrant parents from different origins, their group was decided with a priority given to non-European origin4. Table 5 Life satisfaction determinants: the effect of the ethnic group
Source ESS, round 1, 2 and .3. OLS regression results, the control variables are included in the analysis (see table8 in the appendix)
Models 1.1 and 2.1 provide a test for the ethnic group effect using either all the available data or only rounds 2 and 3. In Model 1.1, only immigrants coming from “other geographical region” (mainly North-America or Australia) have no significant differences in life 4
These cases are very few.
8
satisfaction level than natives. All other ethnic groups seem to have a negative impact on subjective well-being. African first and second generation immigrants are the most dissatisfied with their lives. Their coefficient are significantly different from the Europeans’ one which is still less favourable than natives. Model 2.1 draws the attention to the importance of distinguishing between immigrants coming from Western-Europe and EasternEurope. Only the latters have significant lower life satisfaction levels compared to natives. This model also provides an estimation of the specific effect of a Turkish origin which seems to be the most negative effect among all ethnic groups. According to these findings, the lower life satisfaction level of immigrants and their descendants in Europe seems to be related more to their belonging to an underprivileged minority group than to their specific experience of migration and the acculturation problems it may lead to. Compared to natives, WesternEuropeans and immigrants coming from other geographical areas are the only ones that do not have a lower significant life satisfaction level after controlling for the relevant co-variables. These groups are high skilled workers and do not suffer from ethnic penalties. Members of all other groups, especially African and Turkish, experience lower social status and less favourable objective living conditions. They also seem to be more frequently confronted to prejudice and hostility from the host society. Nevertheless, providing a convincing explanation for this depression in first and second generations’ well-being as related to their belonging to disadvantaged ethnic groups is difficult if only based on their objective unfavourable living conditions. Indeed, an important set of socio-economic variables are controlled for in the models above, at the individual level. A valid justification should therefore come within collective subjective considerations linked to ethnicity. To what extent does the feeling some first and second generation immigrants have that their ethnic group is discriminated against account for their lower life satisfaction levels? Is perceived discrimination the key variable in explaining the relative disadvantage in the life contentment of immigrants? Measuring the effect of discrimination While discrimination is often put forward as a possible explanation of the lower life satisfaction level of ethnic or racial minorities, measuring its specific effect in a regression model poses several problems. Indeed, the introduction of the discrimination variable as a covariable among others in the life satisfaction model may not be relevant because, unlike clear exogenous variables as gender or age, the relation between perceived discrimination and life satisfaction can hardly be considered as directly causal. One may think that other variables may affect both life satisfaction level and the individual answer to the discrimination question. Some of these variables can be observed (ethnic group, age, etc.) and some are unobservable (personality variables). Estimated in a single equation, the effect of perceived discrimination on life satisfaction may thus be fallacious. An attempt to correct this endogeneity bias can be conducted using simultaneous equation models (Maddala 1983 ; Wooldridge 2005). This technique starts estimating a discrimination model and includes afterwards the predicted value of discrimination in the life satisfaction one. In order for the interest coefficient (discrimination) not to be biased, at least one variable should be used in the discrimination equation and excluded from the life satisfaction one (Heckman 1978). This variable is called instrumental. Though the two stage procedure does correct the endogeneity bias, it does not directly provide the exact variance-covariance matrix of the estimates. In order to estimate this matrix correctly and thus give robust significance tests of the coefficients, the bootstrap technique will be used.
9
.2 0
.1
Perceived discrimination
.3
.4
Figure 3 Perceived discrimination among ethnic groups
TU
SA
OT
AF
AS
WE EE
G3
Source: ESS, Round 2 and 3
.3 .2 0
.1
Perceived discrimination
.4
.5
Figure 4 Types of perceived discrimination among ethnic groups
TU
SA
OT
AF
AS
WE EE
G3
Nationality
Ethnicity
Race
Religion
Other discrimination types Source: ESS, Round 2 and 3
In ESS data, the discrimination question was asked in a very general way (and not specifically to ethnic minorities)5. Figure 3 shows that members of all groups (even G3) may 5
The question asked is the following “Would you describe yourself as being a member of group that is discriminated against in this country?”
10
answer positively to this question. The percentage is the most important in the Turkish, African and South-American groups and the lower in the North–American, European and native groups. Providing a closer view on the types of perceived discrimination, figure 4 shows that they differ considerably among groups. Turkish immigrants regard discrimination as against their nationality while the race reason is the most frequently put forward by Africans. SouthAmericans and Asians also very often perceive the discrimination they are confronted to as racial. Racial discrimination is however hardly ever referred to by European immigrants and even less by “non-ethnic” natives6. In order to take into account perceived discrimination as a possible explanation for immigrants’ lower life satisfaction, models 1.1 and 2.1 were re-estimated with the introduction of this variable (models 1.2 and 2.2, table 5). The ethnic group effect looses ground when the discrimination variable is controlled for. Estimates for African, Eastern European and Turkish are consistently lower than those in model 1.1 and 2.1. They are still however significantly negative with regard to the native group (except for Turkish immigrants). Nevertheless, estimation of the discrimination effect in a single equation may not be consistent. The endogenous aspect of this variable requires the use of simultaneous equations with instrumental variables. We used two religion affiliations (Judaism and Islam) to instrument perceived discrimination. Both of them can be supposed to strongly affect perception of discrimination. On the other hand, asserting to belong to Judaism or Islam can hardly be supposed to have a causal effect on life satisfaction level. Judaism and Islam are not used here as a measure of religiosity; Jews and Muslims are compared not only to unreligious persons but also to Catholics, Protestants, Buddhists and other persons that may be more or less religious. Asserting to be affiliated to Islam or Judaism can be understood as being member of a community that may be perceived as discriminated against, no matter one’s level of religious beliefs and practices. 7 Table 6 displays the findings of life satisfaction model with the use of simultaneous equations. Results for the first stage estimation of discrimination are exposed in the appendix. They show that the instrumental variables do have a highly significant effect on perceived discrimination, and are therefore valid. Other variables being controlled for, people who asserted to belong to Judaism or Islam report higher levels of perceived discrimination than the others. As far as life satisfaction is concerned, the main finding is that when discrimination is controlled for in the second stage equation, there is no longer a significant effect of the ethnic group on life satisfaction. The lower life satisfaction of non-European immigrants, especially of Africans, Turkish and Asians thus seems to be directly related to their perception of discrimination. A significant effect remains only for Eastern-European. Members of this group are the only ones whose lower level of individual well-being cannot be attributed to their feelings of being discriminated against. 6
Multiple answers were possible for the discrimination type question. This explains the difference in the global rates in figures 5 and 6. 7 Some studies have shown that religiosity has a positive (small) effect on life satisfaction (Ellison 1991). However, it is not the belonging to a certain religion but rather religious and spiritual beliefs that affects wellbeing. In the present work, we suppose tha religious affiliation is not highly correlated to religiosity; one may assert that he or she is Muslim or Jewish but does not however intensely practice these religions and may even not have any strong religious belief. In ESS data for example, 47.7% of the persons who declared belonging to a religion also declared a degree of religiosity smaller than 5 (in a scale from 0 to 10). 11% of them declared a religiosity score inferior to 2.
11
Table 6 Estimated effect of perceived discrimination (simultaneous equations)
Source ESS, round 1, 2 and .3. OLS regression results, the control variables are included in the analysis (see table s9 and 10in the appendix)
Conclusion Exploiting ESS data from three rounds and several European countries, first we do measure differences in life satisfaction levels between immigrants and their descendants on the one hand and natives on the other hand. These differences do not vanish with immigrants’ length of stay: even after more than twenty years spent in the host country, immigrants still report lower life satisfaction than natives. A major finding concerns the generational effect; despite the fact that they were born and socialised in host countries, when both of their parents are immigrants, second generations seem to be at least as dissatisfied with their life as the first generation members. This result illustrates the specific socio-cultural disruption second generation immigrants experience and the fact, well documented in some qualitative researches, that, more than their parents, they regard their inferior living conditions as fundamentally unfair (Handlin 1966 ; Portes and Rumbaut 2001). While studies on immigrants’ well-being often stress the migration negative psychological effects and the beneficial impact of the assimilation progress, the results of the previous analysis tends to invalidate the idea of a specific “psychological feature” that may explain the lower level of subjective immigrants’ well-being. All immigrants are not relatively dissatisfied with their lives: only the most underprivileged groups have a significantly lower level of happiness. The lasting differences between life satisfaction of ethnic minorities and that of natives, in spite of the control for a great set of variables reflecting inequalities of life conditions, disappear when perceived discrimination is introduced in the analysis. This is all the more true for African, Asian and Turkish immigrants.
12
Appendix Table 7 OLS regression with the assimilation variables
13
Table 8 OLS regression with the ethnic variables
14
Tables 9 and 10 Simultaneous equations a) Regression of discrimination (probit)
15
b) OLS regression of life satisfaction (variance-covariance matrix estimated by bootstrap)
16
REFERENCES Abramson, Harold J. 1994. “Assimilation and pluralism”, p. 150-160 in Stephan. Thernstorn. Harvard encyclopedia of American ethnic groups. Cambridge: Belknap Press of Harvard University Press. Alba, Richard and Victor Nee. 2003. Remaking the American mainstream. Assimilation and contemporary immigration. Cambridge, Massachussetts and London, England: Harvard University Press. Alba, Richard. 1995. “Assimilation's quiet tide”. Public Interest 119:3-18. Baltatescu, Sergiu Marian. 2005. “Subjective well-being of immigrants in Europe and their evaluation of societal conditions. An exploratory study”, p. 128-143 in Lia Pop and Cristina Matiuta. European Identity and Free Movement of Persons in Europe. Oradea: University of Oradea Publishing House. Berry, John W. 2001. “A Psychology of Immigration”. Journal of Social Issues 57:615-631. Berry, John W., Uichol Kim, Thomas Minde, and Doris Mok. 1987. “Comparative Studies of Acculturative Stress”. International Migration Review 21:491-511. Carmichael, Fiona and R. Woods. 2000. “Ethnic Penalties in Unemployment and Occupational Attainment: Evidence for Britain”. International Review of Applied Economics 14:71-98. Castles, Stephen and G. Kosack. 1973. Immigrant Workers and class structures in Western Europe. Oxford: Oxford University Press. Cummins, Robert A. 2000. “Objective and subjective quality of life: an interactive model”. Social Indicators Research 52:55. Ellison, Christopher G. 1991. “Religious Involvement and Subjective Well-Being”. Journal of Health and Social Behavior 32:80-99. Finch, Brian Karl, Bohdan Kolody, and William A. Vega. 2000. “Perceived Discrimination and Depression among Mexican-Origin Adults in California”. Journal of Health and Social Behavior 41:295-313. Frey, Bruno S. and Alois Stutzer. 2002. Happiness and economics. Princeton and Oxford: Princeton University Press. Gordon, Milton M. 1964. Assimilation in American life: the role of race, religion, and national origins. New York: Oxford University Press. Handlin, Oscar. 1951. The uprooted. Boston: Little, Brow and Company. —. 1966. Children of the uprooted. New York: George Braziller. Heath, Anthony and Soojin Yu. 2005. “Explaining Ethnic Minority Disadvantage”, p. 187224 in Anthony F. Heath, John Ermisch, and Duncan Gallie. Understanding social change. Oxford and New York: Oxford University Press. Heckman, James J. 1978. “Dummy Endogenous Variables in a Simultaneous Equation System”. Econometrica 46:931-959. Hughes, Michael and David H. Demo. 1989. “Self-Perceptions of Black Americans: SelfEsteem and Personal Efficacy”. The American Journal of Sociology 95:132-159. Hughes, Michael and Melvin E. Thomas. 1998. “The Continuing Significance of Race Revisited: A Study of Race, Class, and Quality of Life in America, 1972 to 1996”. American Sociological Review 63:785-795. Maddala, G. 1983. “Limited dependent and qualitative variables in econometrics”. Econometric Society Monographs. Cambridge University Press. Neto, Félix. 1995. “Predictors of satisfaction with life among second generation migrants”. Social Indicators Research 35:93-116. —. 2001. “Satisfaction with Life Among Adolescents from Immigrant Families in Portugal”. Journal of Youth and Adolescence 30:53-67.
17
Park, Robert E. 1928. “Human migration and the marginal man”. The American Journal of Sociology 33:881-893. Portes, Alejandro and Ruben G. Rumbaut. 2001. Legacies. The story of the immigrant second generation. Berkeley, CA: University of California Press Portes, Alejandro and Min Zhou. 1993. “The New Second Generation: Segmented Assimilation and Its Variants”. Annals of the American Academy of Political and Social Science 530:74-96. Ramakrishnan, Karthick S. 2004. “Second Generation Immigrants? The "2.5 generation" in the United-States”. Social Science Quarterly 85:380-399. Rooth, Dan-Olof and Jan Ekberg. 2003. “Unemployment and earnings for second generation immigrants in Sweden. Ethnic background and parent composition”. Journal of Population Economics 16:787-814. Rumbaut, Ruben G. 1994. “The Crucible within: Ethnic Identity, Self-Esteem, and Segmented Assimilation among Children of Immigrants”. International Migration Review 28:748-794. —. 1997. “Assimilation and Its Discontents: Between Rhetoric and Reality”. International Migration Review 31:923-960. Safi, Mirna. 2008. “The immigrant Integration Process in France: Inequalities and Segmentation”. Revue française de sociologie 49:3-44. Sellers, Robert M., Cleopatra H. Caldwell, Karen H. Schmeelk-Cone, and Marc A. Zimmerman. 2003. “Racial Identity, Racial Discrimination, Perceived Stress, and Psychological Distress among African American Young Adults”. Journal of Health and Social Behavior 44:302-317. Taylor, John and R. Jay Turner. 2002. “Perceived Discrimination, Social Stress, and Depression in the Transition to Adulthood: Racial Contrasts”. Social Psychology Quarterly 65:213-225. Thomas, Melvin E. and Michael Hughes. 1986. “The Continuing Significance of Race: A Study of Race, Class, and Quality of Life in America, 1972-1985”. American Sociological Review 51:830-841. Vega, William A. and Ruben G. Rumbaut. 1991. “Ethnic Minorities and Mental Health”. Annual Review of Sociology 17:351-383. Werkuyten, Maykel and Shervin Nekuee. 1999. “Subjective Well-Being, Discrimination and Cultural Conflict: Iranians Living in The Netherlands”. Social Indicators Research 47:281-306. Wooldridge, Jeffrey M. 2005. Introductory econometrics: a modern approach: SouthWestern publishing. Zhou, M. 1997. “Segmented assimilation: issues, controversies, and recent research on the New Second Generation”. International Migration Review 31:975-1008.
18
