FAMILY PROBLEMS AND YOUTH UNEMPLOYMENT.

Goede, Martijn de ; Spruijt, Ed ; Maas, Cora 等

ABSTRACT

This study attempted to determine the extent to which family and personal characteristics relate to the employment situation of adolescents. Data were drawn from the Utrecht Study of Adolescent Development (USAD), which investigated, longitudinally, a national sample of Dutch youths aged 12 to 24 years in 1991. Specifically, two waves of a sample of 955 non-school-going respondents between 18 and 27 years old were analyzed. Parental divorce, parental unemployment (only for males), low parental affective involvement, and adolescent relationship problems were related to youth unemployment, but educational career and work commitment were not. For males, parental unemployment demonstrated the strongest correlation with youth unemployment. For females, only variables in the relational domain played a role in explaining unemployment; relationship variables were also important predictors of male unemployment. The results suggest that the family factors included in this study are better predictors of youth unemployme nt than are the classic individual (personal) variables.

In almost all European countries, young people--defined as those under 25 years old--have experienced higher rates of unemployment than have other age groups. The youth unemployment rate in 1993 exceeded 30% in Italy, Spain, and Finland, compared with 13.3% in the United States, 18% in Canada, and only 5% in Japan. Only the few European countries with a traditionally strong apprenticeship system--Austria, Germany, and Switzerland--succeeded in maintaining low youth unemployment rates (Organisation of Economic Cooperation and Development, 1994).

Youth unemployment is recognized as a serious societal problem, but it also has other repercussions. Because the work environment provides opportunities for learning, showing initiative, and developing social contacts and self-reliance (Warr et al., 1985; Warr, 1987), unemployment can be expected to have a negative impact on the growth, and even mental health, of the individual. Adolescents are especially vulnerable to the negative effects of unemployment.

Adolescence is an important life stage in terms of psychosocial development, particularly identity formation. Further, it usually involves a series of status passages. Young people may have to make the transition from high school to college, or from school to work, or from the parental home to a household of their own, or all of these (de Goede, de Klaver, Van Ophem, Verhaar, & de Vries, 1996). A delay in the transition from school to work may start a process of social exclusion that, for some adolescents, has long-term consequences (Te Grotenhuis, 1994).

Individual (personal) variables, such as education, sex, and age, are considered important predictors of youth unemployment. Three aspects of the educational career seem to be relevant to a person's position in the labor market: the level of education, the specialization, and the diploma obtained. Continued schooling is frequently mentioned as a solution to unemployment. It is generally believed that greater education (vocational or university) will lead to better job opportunities. Beker and Merens (1994), however, reported that the lowest unemployment rates were among adolescents who had obtained a certificate of lower vocational training or a lower general secondary education. Adolescents with intermediate vocational training, higher general secondary education, or a pre-university education had the next best rates, followed by adolescents with higher vocational training. University students had a relatively bad position in the labor market, a phenomenon that is fairly new. Nevertheless, adolescents enteri ng the labor market without any certificate or diploma have great problems in finding a job. Furthermore, many courses of study with a majority of female students (exceptions include medicine and nursing) lead to a relatively weak position in the labor market (Rapportage Arbeidsmarkt, 1991).

There are also indications that family background variables are important predictors of educational career and subsequent youth unemployment. Different aspects of family life seem to be important in this regard: parental divorce, parental unemployment, and the degree of affective involvement of parents.

In terms of parental divorce, children from divorced families generally show poorer patterns of adjustment than do those from intact families. For example, divorce has been shown to place children at greater risk for delinquency and substance abuse, early sexual activity, teenage pregnancy, school failure, and emotional problems (Amato & Keith, 1991; Emery, 1988; McLanahan & Booth, 1989), with often long-lasting repercussions. Several studies have found that adults who experienced parental divorce as children tend to have lower socioeconomic attainment, greater marital instability, and poorer psychological adjustment when compared with those who grew up in continuously intact families (Amato & Keith, 1991; McLanahan & Sandefur, 1994). Parental divorce might have negative effects on children's attitudes and behavior (e.g., self-presentation), with implications for getting and keeping a job.

Similar to divorce, parental unemployment might affect some children's attitudes and behavior (Derks, Elchardus, Glorieux, & Pelleriaux, 1996). Being out of work may lead to financial hardship, parental depression and anxiety, and uncertainty about the future. It is unlikely that children would be unaffected by such circumstances (Madge, 1983).

In general, research on child development has shown that children who have good family relationships experience fewer problems. In Boss's (1987) models of family functioning, parental affective involvement plays a significant role in child development. In families coping with problems, it is especially important that children are assured of emotional support. Lack of parental affective involvement can have long-term negative effects on children's development, such as in the areas of self-esteem and self-consciousness, which are relevant for the domain of employment.

The present study did not deal with the macroeconomic causes of the societal problem of youth unemployment. Instead, it dealt with the aforementioned problems in the family of origin-parental divorce, parental unemployment, and lack of parental affective involvement as predictors of youth unemployment. Moreover, it also investigated individual variables that are often used in explaining youth unemployment: sex, age, education, work commitment, and parental socioeconomic status (de Goede & Maassen, 1986). Thus, a comparison of the explanatory power of individual and family variables was possible.

RESEARCH QUESTIONS AND HYPOTHESES

The aim of this study was to gain greater insight into the relationship of family and personal characteristics to the work situation of adolescents. The following research questions were formulated: To what extent do parental (family) characteristics relate to the employment situation of adolescents? To what extent do individual characteristics relate to the employment situation of adolescents? Are these relationships different for males and females? Several hypotheses were then investigated.

Hypothesis 1

Many children and adolescents are confronted with parental divorce, a major life disruption. One common assumption is that children often experience severe and long-term postdivorce adjustment problems. Studies using small clinical samples have tended to support this assumption (e.g., Wallerstein & Blakeslee, 1990). Most larger and more representative studies have found moderate and mostly short-term effects, both for adults and for children (Demo & Ganong, 1994). Amato and Keith (1991) conducted a meta-analysis of 37 studies (involving over 81,000 individuals) dealing with the long-term consequences of parental divorce. Effect sizes were calculated for 15 outcome variables. It was concluded that parental divorce has significant negative effects on children's later well-being, specifically in terms of teenage pregnancy, teenage marriage, social well-being, the quality of marriage, divorce, and physical health.

Bosman and Louwes (1988) reported on the effects of divorce on the school careers of children. They concluded that the family transition in the case of divorce is often characterized by an accumulation of problems that seem to harm children's chances of school success. Dronkers (1992) found significant differences in the school careers of children from two-parent and one-parent families; the latter showed a lower level of achievement.

Parental divorce seems to adversely affect not only educational achievement, but also the personal characteristics of children. These include social skills, flexibility, and self-presentation, which are important in terms of employment prospects (Buwalda & de Vries, 1994).

In short, children of divorced parents appear to face negative consequences in the domains of health, education, and general personal characteristics, all of which are relevant to future employment. Parental divorce, in general, may limit children's later chances of getting or keeping a job. This led to hypothesis 1: There is a positive relationship between parental divorce and youth unemployment.

Hypothesis 2

Regarding child and adolescent development, Jahoda, Lazarsfeld, and Zeisel (1933) were the first to emphasize the psychological significance of economic deprivation as a consequence of parental unemployment. Elder and Caspi (1988) found that adolescents' well-being and conduct were affected by financial problems, with increased strain in family relationships as an intervening variable. Family processes play a central role in linking economic problems to self-derogation, tendency toward transgression, and other aspects of maladjustment (Silbereisen, Walper, & Albrecht, 1990). Maladjustment reduces the chances of getting and holding a job.

Derks et al. (1996) stressed that the effects of parental unemployment on the employment situation of children are not direct, but are mediated by educational performance. Derks et al. found that the probability of school failure is greater for children of unemployed fathers, with obvious consequences for labor market position. Unemployment may thus be passed on from one generation to the next through the mechanism of educational career.

The consequences for children may be particularly serious when both parents have had less education and are unemployed. Te Grotenhuis (1994) indicated that unemployed parents have smaller and less varied social networks than do employed parents, diminishing the availability of, and benefits accruing from, "social capital" (Bourdieu, 1996). In other words, they have less access to employment information or opportunities derived via people in their environment.

Thus, parental unemployment weakens children's position in the labor market. In a way, long-term unemployment is self-reinforcing, even over several generations. This led to hypothesis 2: There is a positive relationship between parental unemployment and youth unemployment.

Hypothesis 3

Good family relations are, in general, important for the development of mature attitudes and behavior. According to the McMaster model of family functioning, parents need to be affectively involved with their children (Olson & McCubbin, 1983). Several studies (Maccoby & Martin, 1983; Steinberg, Mounts, & Dornbusch, 1991) have shown that adolescents who characterize their parents as neglectful are consistently compromised in the areas of competence, self-perception, and misbehavior, and they tend to be psychologically distressed. Additional research (McBride & Bell-Scott, 1994) has suggested that lack of parental affective involvement increases the likelihood of a negative outcome for adolescents, such as immature attitudes and maladjusted behavior. These, in turn, decrease adolescents' chances of getting and keeping a job. This led to hypothesis 3: There is a negative relationship between parental affective involvement and youth unemployment.

Hypothesis 4

In the area of personal relationships, young people in the Netherlands, on average, start their first steady opposite-sex relationship around age 16. The dominant relationship pattern is to live with a boyfriend or girlfriend, often followed by marriage after a few years. About half of these initially cohabitational relationships end in divorce.

Adolescents who have relationship setbacks will likely experience psychological stress and are susceptible to diminished self-esteem (Gerstel, 1990). Self-presentation and social skills may then deteriorate, leading to poorer chances of getting and keeping a job. This led to hypothesis 4: There is a positive relationship between adolescent relationship problems and youth unemployment.

Hypothesis 5

Bynner (1996) concluded that young people with very low levels of basic and work-related skills have little chance of getting a job and, if they get one, of keeping it. Studies by Wilson (1987) and Te Grotenhuis (1994) clearly underscored the importance of education: no education, no job prospects (see also de Goede et al., 1996). Thus, in line with hypothesis 2, youths with low levels of education are in a weak position in the labor market and have a greater chance of being and staying unemployed. This led to hypothesis 5: There is a negative relationship between educational career and youth unemployment.

Hypothesis 6

In principle, society tries to foster a work ethic. People are expected to be economically independent as much as possible and to contribute to society. A low work commitment is associated with passivity regarding the job search and application process, and a critical attitude toward the jobs that are available (Raaijmakers, 1986; de Goede & Maassen, 1986). Along these lines, adolescents with a strong work commitment will be motivated to get a stable job. If they become unemployed, they will search actively for work and have a less critical attitude toward what is available. This led to hypothesis 6: There is a negative relationship between the degree of work commitment and youth unemployment.

Hypotheses 7a and 7b

In general, females' identity is tied strongly to their social relations, while males' identity is tied strongly to their occupation (Meeus & 't Hart, 1993). Thus, females will be more sensitive to relationship problems, and males will be more affected by difficulties in the vocational domain (de Goede & Hustinx, 1993). This led to hypothesis 7a: Problems in the vocational domain have a stronger negative association with the employment of males than females. Similarly, it led to hypothesis 7b: Problems in the relationship domain have a stronger negative association with the employment of females than males.

METHOD

Subjects

The subjects for this study were participating in a longitudinal research project, the Utrecht Study of Adolescent Development (USAD; see Meens & 't Hart, 1993). Two waves of this project, 1991 and 1994, were available. A national sample of Dutch youths, aged 18 to 24 years in 1991, were interviewed at home and also completed an extensive questionnaire that they returned later. In addition, one parent was interviewed. In some cases, a second or even a third youth in the family was included in the sample. This manner of sampling offered an opportunity for multilevel data analysis.

Those with missing data were dropped from the analysis, resulting in a sample of 955 subjects (41% male and 59% female). Their average age was 23.37 years. Nearly 40% had experienced unemployment, and 20% had experienced serious financial problems. At the time of wave 2 (1994), 10.5% were unemployed. Fifteen percent had experienced parental divorce (10% for less than three years and 5% for more than three years). See Table 1 for additional background data.

Measures

Youth unemployment (dependent variable). This variable was assessed in the first and second waves. The 1991 score was combined with the 1994 score, using a scale ranging from 0 (did not experience unemployment) to 4 (experienced unemployment and was deeply affected by it).

Parental factors. Information on divorce was obtained in the first and second waves (1 = no, 2 = more than three years, 3 = less than three years).

Parental unemployment was assessed in the first and second waves for fathers and mothers separately. The 1991 score was combined with the 1994 score, using a scale ranging from 0 (did not experience unemployment) to 4 (experienced unemployment and was deeply affected by it). The father's score and the mother's score were then combined, using a scale ranging from 0 (not experienced by either parent) to 8 (experienced by both and deeply affected by it).

Parental affective involvement was measured through a combination of two scales (i.e., father's affective involvement and mother's affective involvement). Both consisted of 10 items; for example: "My father (mother) talked to me with warmth and in a friendly way"; "My father (mother) did appear to understand my problems and cares"; and "My father (mother) appreciated that I made my own decisions." Possible answers were 0 = not true, 1 = more or less not true, 2 = more or less true, and 3 = true. The total score for this combined scale had a range of 0 (weak) to 60 (strong). Cronbach's alpha was .83.

One indicator of socioeconomic status was family income per week (1 = the equivalent of 150 U.S. dollars or less to 12 = 600 U.S. dollars or more). Other indicators were education of father and education of mother (1 = low to 7 = high). Parental education was a combination of both scores, with a range from 2 (low) to 14 (high).

Individual factors. Problems in personal relationships (questions dealt with splitting up after courtship/cohabitation and/or divorce) were assessed in the first and second waves. The 1991 score was combined with the 1994 score, using a scale ranging from 0 (did not experience the particular relationship problem) to 4 (experienced it and was deeply affected). The relationship problems variable consisted of overall difficulties, ranging from 0 (no experience) to 8 (two or more such experiences and deeply affected by them).

There were several indicators of educational career. One was number of years attending primary school (5 to 8 years), with more than 6 years implying repetition of one or more classes/years (i.e., insufficient progress during the school year). Another was school career, which was based on the difference between the type of secondary education chosen directly after primary school and the type of school attended (in wave 2, 1994), or the highest type of education completed for those with a paying job (-6 downward move to +6 upward move in school career). Number of diplomas (0 = no diploma, 1 = 1, to 4 = 4 or more) was also an indicator of educational career. Finally, there was level of education (1 = low to 7 = high).

Work commitment--the extent to which people find having or getting a paying job important--was assessed. This measure consisted of 11 items; for example: "To make something out of life, one needs to find steady employment" and "Making a career for myself is important to me." The scale ranged from 11 (low) to 55 (high). Cronbach's alpha was .88.

Additional personal factors were sex (1 = boy, 2 = girl); age (16-27 years); income (net amount of money) per month (1 = the equivalent of 325 U.S. dollars or less to 10 = 1,250 U.S. dollars or more); and serious financial problems (being in debt). This latter variable was assessed in the first and second waves. The 1991 score was combined with the 1994 score, using a scale ranging from 0 (did not experience serious financial problems) to 4 (experienced serious financial problems and deeply affected by them).

Analysis

The USAD data included observations of one or more children within a family, as well as observations of family context. This seemed to require the hierarchical linear or multilevel model of analysis (Hox, 1994; Goldstein, 1995; Bryk & Raudenbush, 1992). The MLn computer program (Rasbash & Woodhouse, 1995) was consequently used. It enabled the intrafamily correlation to be calculated. This correlation is an estimate of the percentage of variance at the family level. Within-family observations are more alike than between-families observations if the variance at the family level differs significantly from zero. This implies a violation of a crucial assumption of ordinary regression analysis (i.e., the independence of observations). However, for the 955 youths included in this study, the intrafamily correlation (r) was .00. Two youths in the same family were present in only 63 cases. The multilevel model was therefore unnecessary and, instead, OLS regression was used.

First, bivariate correlations between the employment situation of the youths and all the other variables were calculated. Second, variables that correlated significantly with employment situation were selected for the regression analysis. Youth income and serious financial problems, which can be considered consequences of the employment situation, were not included in the regression analysis. The regression analysis was also executed separately for males and females because of the assumed differences between them (see hypotheses 7a and 7b).

RESULTS

The results of the OLS regression analysis (see Table 2) were examined with respect to the hypotheses.

It was hypothesized that parental divorce would be positively related to youth unemployment. In fact, greater divorce was significantly associated with higher unemployment (beta = .l0,p [less than] .01). Thus, hypothesis 1 was confirmed.

Hypothesis 2 was rejected. Parental unemployment was not significantly related to youth unemployment for the total sample (beta .07, p = .07).

As expected, parental affective involvement correlated negatively with youth unemployment (hypothesis 3). The weaker the affective involvement of parents, the more often youths were confronted with unemployment (beta = - .11, p = .004).

Hypothesis 4 stated that youths' relationship problems would positively correlate with youth unemployment. This was corroborated (beta = .10,p [less than] .01).

Hypothesis 5 stated that there would be a negative correlation between educational career and youth unemployment. There were no significant betas for any of the indicators of education. Therefore, this hypothesis was rejected.

Work commitment was expected to correlate negatively with youth unemployment. The assumption was that unemployed youths would have a weaker work commitment as compared with those with a job. There appeared to be no correlation between these variables and therefore work commitment was not included in the regression analysis. Thus, hypothesis 6 must be rejected.

Sex differences were also examined. For example, parental unemployment was positively correlated with youth unemployment for males (beta = .18, p [less than] .001) but not for females (beta = - .01). The difference between these betas was significant (p = .012; see Kamaian & Raudenbush, 1996). In hypothesis 7a, it was predicted that problems in the vocational domain would have a stronger negative association with males' employment than females' employment. This hypothesis was corroborated. Hypothesis 7b stated that problems in the relationship domain would have a stronger negative association with females' employment than males' employment. For females, only variables in the relationship domain, namely parental affective involvement (beta = - .10, p [less than] .05) and relationship problems (beta = .09, p [less than] .05), were significant, in line with the hypothesis. However, these relationship variables were also significant or approached statistical significance for males. These findings provide a parti al confirmation of hypothesis 7b.

DISCUSSION AND CONCLUSIONS

The findings offer insight into the separate relationships of parental and personal characteristics to youth unemployment. Various family factors, such as parental divorce, parental unemployment, and parental affective involvement were examined, all of which have attracted little attention in the literature on youth unemployment. An important aim of this study was to compare these family factors with some classic variables in their ability to explain youth unemployment.

Most of the hypotheses were confirmed. Parental divorce, parental unemployment (only for males), parental affective involvement, and youth relationship problems played a significant role in explaining youth unemployment.

Hypotheses on the relationship of unemployment to two individual variables, educational career and work commitment, had to be rejected. These variables did not seem to play a significant role in youth unemployment. Clearly, youths have to have some education and a commitment to get and keep a job, but these basic conditions do not guarantee employment.

For males, parental unemployment demonstrated the strongest correlation with youth unemployment. Therefore, we can speak of an intergenerational transmission of male unemployment. This is in agreement with previous findings (see, for example, Derks et al., 1996).

For females, only variables in the relationship domain played a role in explaining unemployment (one of these variables also was a significant predictor for males, and the other approached statistical significance). These results agree with the literature, in that female identity tends to be tied to their social relations, while male identity is based more or less equally on their relationships and on vocational issues (Meeus & 't Hart, 1993).

The results imply that family factors are more important as predictors of youth unemployment than are the classic variables, and highlight the role of family socialization in the behavior of adult children. This interpretation is consistent with several qualitative descriptions of the problems that adult children of divorce experience (Wallerstein, Corbin, & Lewis 1989). Through the process of socialization, parental divorce and low parental affective involvement may increase the likelihood that children will not learn adaptive interpersonal skills, such as how to reach a compromise and communicate effectively (see Amato, 1996). This, in turn, handicaps their job prospects. Of course, there may be alternative explanations. Parents with personality deficits, for example, may pass these characteristics to their children (e.g., through their child-rearing methods), increasing the risk of unemployment and/or divorce (Amato, 1996).

In sum, this study has yielded a number of findings that contribute to our understanding of the connection between some basic family problems and youth unemployment. However, more research is needed to test alternative explanations and to examine the complex interactional effects of family problems on adolescents. In particular, more attention has to be paid to the intervening variables between family problems and youth unemployment.

Finally, it should be noted that problems in the family of origin increase the risk of unemployment, but the majority of adult children of divorced, of unemployed, and of affectively remote parents do get and continue to hold jobs. Moreover, although several of the correlations in this empirical study were statistically significant, they were not very large. Parental troubles affect, but certainly do not wholly determine, their children's employment situation.

This research was supported by a grant from the Dutch Organization for Scientific Research (NWO) to the Utrecht Study on Adolescent Development (USAD) 1991-1997.

Ed Spruijt, Department of Child and Youth Studies, Utrecht University.

Cora Maas, Department of Methodology and Statistics, Utrecht University.

Vincent Duindam, Department of Communication and Welfare, Utrecht University.

REFERENCES

Amato, P. R. (1991). The "child of divorce" as a person prototype: Bias in the recall of information about children in divorced families. Journal of Marriage and the Family, 53, 59-70.

Amato, P. R. (1996). Explaining the intergenerational transmission of divorce. Journal of Marriage and the Family, 58, 628-640.

Amato, P. R., & Keith, B. (1991). Parental divorce and adult well-being: A meta-analysis. Journal of Marriage and the Family, 53, 43-58.

Beker, M., & Merens, J.G.F. (1994). Rapportage jeugd, 1994 [Report on youth, 1994]. Rijswijk: Sociaal en Cultureel Planbureau.

Bird, G., & Melville, K. (1994). Families and intimate relationships. New York: McGraw-Hill.

Bosman, R., & Louwes, W. (1988). Eenouder- en tweeoudergezinnen en school-loopbanen [Single-parent and two-parent families and school careers]. Mens en Maatschappij, 63, 5-23.

Boss, P. (1987). Family stress. In M. B. Sussman & S. K. Steinmetz (Eds.), Handbook of marriage and the family (pp. 695-723). New York: Plenum.

Bourdieu, P. (1996). Opstellen over smaak, habitus en het veldbegrip [Essays on taste, habit and the field concept]. Amsterdam: Van Gennep.

Bryk, A. S., & Raudenbush, S. W. (1992). Hierarchical linear models: Applications and data analysis methods. Newbury Park, CA: Sage.

Buwalda, G. M., & de Vries, A. (1994). Schooling a panacea? In C.H.A. Verhaar & P. M. de Kiaver (Eds.), The functioning of economy and labour market in a peripheral region - The case of Friesland (pp. 184-209). Leeuwarden: Fryske Akademy.

Bynner, J. (1996). Resisting youth unemployment: The role of education and training. In M.P.M. de Goede, P. M. de Klaver, J.A.C. Van Ophem, C.H.A. Verhaar, & A. de Vries (Eds.), Youth: Unemployment, identity and policy (pp. 13-29). Aldershot: Avebury.

De Goede, M. P. M., de Klaver, P., Van Ophem, J., Verhaar, K., & de Vries, A. (1996). Questions on youth unemployment, identity formation and policy. In M. P. M. de Goede, P. M. de Kiaver, J. A. C. Van Ophem, C. H. A. Verhaar, & A. de Vries (Eds.), Youth: Unemployment, identity and policy (pp. 1-12). Aldershot: Avebury.

De Goede, M. P. M., & Hustinx, P. (1993). School en beroep [School and profession]. In W. Meeus & H. 't Hart (Eds.), Jongeren in Nederland [Youth in the Netherlands] (pp. 79-105). Amersfoort: Academische Uitgeverij.

De Goede, M. P. M., & Maassen, G. H. (1986). Werkloos-zijn: Hoe beleven jongeren dat? In M. Matthijssen, W. Meeus, & F. Van Wel (Eds.), Beelden van jeugd: Leefwereld, beleid, onderzoek [Images of youth: Life world, policy, research] (pp. 156-170). Groningen: Wolters Nordhoff.

Demo, D. H., & Ganong, L. H. (1994). Divorce. In P. C. McKenry & S. J. Price (Eds.), Families and change: Coping with stressful events (pp. 197-218). Thousand Oaks, CA: Sage.

Derks, A., Elchardus, M., Glorieux, I., & Pelleriaux, K. (1996). The effect of fathers' unemployment on school careers and values of children. In M. P. M. de Goede, P. M. de Klaver, J. A. C. Van Ophem, C. H. A. Verhaar, & A. de Vries (Eds.), Youth: Unemployment, identity and policy (pp. 87-106). Aldershot: Avebury.

Dronkers, J. (1992). Zullen wij voor de kinderen bij elkaar blijven? De veranderende effecten van eenoudergezinnen op de schoolloopbanen van de kinderen [Changing effects of single-parent families on the school careers of children]. Mens en Maatschappij [Man and Society], 67, 23-44.

Elder, G. H., Jr., & Caspi, A. (1988). Human development and social change: An emerging perspective on the life course. In N. Bolger, A. Caspi, G. Downes, & M. Moorehouse (Eds.), Persons in context: Developmental processes (pp. 77-113). Cambridge, MA: Cambridge University Press.

Emery, R. E. (1988). Marriage, divorce, and children's adjustment. Newbury Park, CA: Sage.

Gerstel, N. (1990). Divorce and stigma. In C. Carlson (Ed.), Perspective on the family: History, class, and feminism (pp. 460-478). Belmont, CA: Wadsworth.

Goldstein, H. (1995). Multilevel statistical models (2nd ed.). London: Edward Arnold.

Hox, J. J. (1994). Applied multilevel analysis. Amsterdam: TT-Publikaties.

Jahoda, M., Lazarsfeld, P. F., & Zeisel, H. (1933). Die arbeitslosen von Marienthal: Ein soziografischer Versuch [The unemployed of Marienthal: A sociography attempt]. Frankfurt: Suhrkamp.

Kamaian, H. A., & Raudenbush, S. W. (1996). A multivariate mixed linear model for meta-analysis. Psychological Methods, 1, 227-235.

Maccoby, E., & Martin, J. (1983). Socialization in the context of the family: Parent-child interaction. In E. M. Hetherington (Ed.), Handbook of child psychology: Vol. 4. Socialization, personality, and social development. New York: Wiley.

Madge, N. (1983). Unemployment and its effects on children. Journal of Child Psychology and Psychiatry and Allied Disciplines, 2, 311-319.

McBride, M. V., & Bell-Scott, P. (1994). Dealing with adolescent children. In P. C. McKenry & S. J. Price (Eds.), Families and change: Coping with stressful events. Thousand Oaks, CA: Sage.

McLanahan, S., & Booth, K. (1989). Mother-only families: Problems, prospects, and politics. Journal of Marriage and the Family, 51, 557-580.

McLanahan, S., & Sandefur, G. (1994). Growing up with a single parent. Cambridge, MA: Harvard University Press.

Meeus, W., & 't Hart, H. (Eds.). (1993). Jongeren in Nederland [Youth in the Netherlands]. Amersfoort: Academisch Uitgeverij.

Organisation of Economic Cooperation and Development. (1994). The OECD jobs study: Facts, analysis, strategies. Paris: OECD.

Olson, D. M., & McCubbin, H. I. (1983). Families: What makes them work. Beverly Hills, CA: Sage.

Raaijmakers, Q. (1986). Arbeidsethos bij schoolverlaters. In M. Matthijssen, W. Meeus, & F. Van Wel (Eds.), Beelden van Jeugd: Leefwereld, beleid, onderzoek [Images of youth: Life world, policy, research] (pp. 141-155). Groningen: Wolters Noordhoff.

Rapportage Arbeidsmarkt 1991 [Report on labor market 1991]. (1991). Den Haag: Ministerie van Sociale Zaken en Werkgelegenheid.

Rasbash, Y., & Woodhouse, G. (1995). MLn command reference. London: Multilevel Models Project, University of London.

Silbereisen, R. K., Walper, S., & Albrecht, H. T. (1990). Families experiencing income loss and economic hardship: Antecedents of adolescents' problem behavior. In V. C. Leloyd & C. Flanagan (Eds.), Risk and protective factors in children and adolescents' response to economic crises and deprivation. San Francisco: Jossey-Bass.

Steinberg, L., Mounts, N. S., & Dornbusch, S. (1991). Authoritative parenting and adolescent adjustment across various ecological niches. Journal of Research on Adolescence, 1, 19-36.

Te Grotenhuis, H. (1994). De erfelijkheid van werkloosheid. In C. Brinkgreve, B. Kruithof, B. van Stolk, & G. de Vries (Eds.), Overdragen en eigen maken; Over sociale erfenissen. [Transfer and acquire: About social inheritances.] Amsterdams Sociologisch Tijdschrift [Amsterdam Sociological Review], 21, 61-85.

Wallerstein, J. S., & Blakeslee, S. (1990). Second chances. New York: Ticknor and Fields.

Wallerstein, J. S., Corbin, S. B., & Lewis, J. M. (1988). Children of divorce: A 10-year study. In E. M. Hetherington & J. D. Arateh (Eds.), Impact of divorce, single parenting, and stepparenting on children. Hillsdale, NJ: Lawrence Erlbaum Associates.

Warr, P. B., Banks, M. H., & Ullah, P. (1985). The experience of unemployment among black and white urban teenagers. British Journal of Psychology, 76, 75-87.

Warr, P. J. (1987). Work, unemployment and mental health. Oxford: Clarendon Press.

Wilson, W. J. (1987). The truly disadvantaged: The inner city, the underclass, and public policy. Chicago: University of Chicago Press.