We analyze the impact on schooling outcomes of growing up in a non-intact family in Germany. We find that this experience is associated with worse outcomes according to estimates from models that do not control for possible correlations between common unobserved determinants of family structure and educational performance. Evidence of adverse effects emerges also when endogeneity is accounted for. In such cases, however, the point estimates are typically smaller, and their confidence intervals are large enough to include zero, particularly for individuals who grew up in Western Germany.
This is a preview of subscription content, access via your institution.
Buy single article
Instant access to the full article PDF.
Price excludes VAT (USA)
Tax calculation will be finalised during checkout.
Two recent examples of a similar multimethod approach are Levine and Zimmerman (2005) and Currie and Tekin (2006). They use a variety of statistical methods to measure the effects of a child’s exposure to welfare benefit receipt on developmental outcomes, and child maltreatment on crime, respectively.
A technical Appendix, where we discuss in detail the identification problems common to all studies of the relationship between childhood family structure and child outcomes and where we outline the empirical strategies underlying each of the econometric methods used in our analysis, is available at: http://www.diw-berlin.de/soep/26652.html?uid=tsiedler.
Arguably, death is not a completely exogenous event and that some of the factors that determine it may also play a role in determining the adult outcomes of children. For example, parents’ educational attainment and socioeconomic status are likely to be powerful drivers of mortality (e.g., Marmot et al. 1984; Elo and Preston 1996). These, in turn, might be correlated with parental smoking, alcohol consumption, and physical activity levels, which may significantly raise the probability of death. Lantz et al. (1998), however, find that such health behaviours explain no more than a modest 12% of the predictive effect of income on mortality.
The SOEP is documented at http://www.diw.de/english/sop/service/index.html.
Sample membership refers to the location when the household was originally sampled, and not current location, because of subsequent mobility between the former East Germany and West Germany. Foreign children, other than those from Guestworker families, were excluded from the analysis due to small sample sizes: nine children from the West German sample and one from the East German sample were dropped.
Father-only families were excluded from the sample: only 75 children (or 2% of individuals in our final sample) were dropped.
See Dustmann (2004) for further details.
This discussion refers to West Germany. After reunification, East Germany adopted the educational system of West Germany (Jeschek 2000). Before 1990, the GDR had a similar school system, albeit with some differences in the length of the various secondary school tracks (e.g., completion of the Gymnasium track required 8 rather than 9 years). Such differences are inconsequential for the measurement of our dependent variables. They only marginally affect our measures of maternal education, but this does not drive any of the differences in results for the East German and West German samples. They are irrelevant for the estimation of mother FE models.
We also analyzed the probability that a child repeated a grade (i.e., whether he/she was ever held back in school) during primary school years. The results of this analysis were similar to those found for test scores and are therefore not reported.
Information on secondary school track at 14 was obtained from parents. For this outcome, we restricted our analysis to children who were enrolled at one of the three main types of secondary school (Hauptschule, Realschule, and Gymnasium).
Ginther and Pollak (2004) distinguished children reared in “blended” families—stepchildren and their half-siblings who are the biological children of both parents—from children reared in traditional intact families as well as from children reared in other family structures. Because of small sample sizes, however, our measures cannot make this distinction.
For children born outside of a partnership before 1983 and for the mother’s marital histories prior to 1983, we cannot know exactly whether the mother cohabited with or married the biological father. For the 255 children (less than 9% of the individuals in the three samples pooled) whose mother partnered within 1 year, we assumed that she moved in with the biological father. Similarly, divorces before 1983 refer to breakdowns of legal marriages only, whereas during the panel years, they cover both legal marriage and cohabitation disruptions.
In earlier work (Francesconi et al. 2005), however, we defined childhood as spanning years 0–16 rather than 0–10 as here. Changing the definition did not change any of our substantive conclusions.
We also experimented with another measure that further distinguished mothers who repartnered after divorce or husband’s death from mothers who did not. We do not report the estimates for such a measure because of the small size of the samples on which this analysis was performed, especially for the East German and Guestworker samples. But the results are qualitatively similar to those shown below.
The greater proportion of highly educated mothers in the East German sample reflects the different educational systems operating in the two Germanies before unification (Frick 2007). This, in turn, is captured by the categorization that we use in the empirical analysis. In particular, the maternal education variable has four categories, in ascending order: general secondary school qualification or less, intermediate school qualification, Abitur, technical college or university degree. To simplify cross-sample comparisons, we used the same broad categories for each sample, though qualifications in the former FRG were different from those in the former GDR, and qualifications in Germany differ from those obtained abroad by mothers in the Guestworker sample. Using an alternative categorization of educational qualifications for mothers, i.e., distinguishing between mothers with engineering and technical college degrees from mothers with university degrees, did not change our key results presented in the next section.
We also investigated the extent to which members of the sibling subsamples and main samples had the same characteristics. Considering the control variables reported in Table 3 (with the obvious exception of only children) and the dependent variables, we found no significant difference between the sibling sample and the main sample at conventional significance levels for all variables except for number of brothers and sisters.
We also performed this analysis using local linear regression matching models and Chamberlain conditional logit models. Because results obtained were similar to those shown in Table 4, they are not reported.
We performed a number of tests to check how the matching on the estimated propensity score successfully balanced the distribution of covariates between treatment and control group. First, we computed standard two-sample mean t tests between treatment and comparison group for all covariates used after matching. Second, we calculated the Pseudo-R 2 from probit estimation of the conditional probability of growing up in a non-intact family both before and after matching (Sianesi 2004). Third, we calculated standardized biases between the treated and control groups before and after matching (Rosenbaum and Rubin 1985). The various test statistics indicated that matching was successful.
In other regressions (not reported for brevity), we also controlled for childhood family income and maternal employment in the estimation of the FE model. Their inclusion did not change our main results.
An argument that differential benefit treatments after parental divorce and after death of a parent dilute identification power is weak. In Germany, if a father dies during childhood, his widow is entitled to receive up to 40% of his earnings. After parental divorce, there is little scope for divorcing parents to agree on bilateral financial transfers in the shadow of the law. In general, the courts establish the exact level of child maintenance, which is typically determined on the basis of the child’s age and the noncustodial parent’s income. There are also guidelines that guarantee a minimum child transfer, independently of the economic circumstances of the noncustodial parent. Overall, the two types of transfers are of comparable levels.
Estimates for “Father died” from the FE model are not presented because of the small number of households with within-family variation (see discussion in Section 3.4 and Table 2 notes). Additional analysis in which we distinguish this treatment reveals that having experienced a parental divorce reduces the chances of achieving Abitur or higher qualifications by almost 19 percentage points, but this effect is not statistically significant at conventional levels (p-value = 0.107). Similar considerations hold when the comparison is performed with respect to individuals who were born to unmarried mothers.
As a robustness check, we reestimated the model for this sample also including a set of dummy variables for mothers’ and fathers’ nationality. The estimates on the family structure variables were very similar to those reported in Table 5, while the nationality dummies were jointly statistically insignificant.
Before 1990, migration between the former GDR and FRG was virtually inexistent. Since then, migration is allowed, but there is one uniform legal family code applied to the whole of Germany. Hence, our results are unlikely to suffer from selective migration bias whereby migration decisions are related to divorce regimes.
Excluding parents who divorced in 1976 and 1977 from the West German sample meant dropping 13 observations, i.e., 4% of all divorced mothers in the sample (or 0.5% of all mothers). Importantly, for the estimation of Gymnasium attendance, 1976 and 1977 are included as pre-reform years for individuals from the former GDR; otherwise the control group would not have information on the pre-reform period.
The different timing of the region variables is because the SOEP does not ask respondents about housing and residential location prior to their joining the panel.
For brevity, the estimates are not shown, but are available from the authors.
The smaller size of these subsamples reduced the precision of some of such estimates, however.
Amato PR (1993) Children’s adjustment to divorce: theories, hypotheses, and empirical support. J Marriage Fam 55(1):23–38
Behrman JR, Rosenzweig MR, Taubman P (1994) Endowments and the allocation of schooling in the family and in the marriage market: the twins experiment. J Polit Econ 102(6):1131–1174
Biblarz TJ, Gottainer G (2000) Family structure and children’s success: a comparison of widowed and divorced single–mother families. J Marriage Fam 62(2):533–548
Björklund A, Sundström M (2006) Parental separation and children’s educational attainment: a sibling analysis on Swedish register data. Economica 73(292):605–624
Björklund A, Ginther DK, Sundström M (2007) Family structure and child outcomes in the United States and Sweden. J Popul Econ 20(1):183–201
Bohrhardt R (2000) Familienstruktur und Bildungserfolg. Zeitschrift für Erziehungswissenschaft 3(3):189–207
Cameron SV, Heckman JJ (2001) The dynamics of educational attainment for black, Hispanic, and white males. J Polit Econ 109(3):455–499
Case A, Lin IF, McLanahan S (2001) Educational attainment of siblings in stepfamilies. Evol Hum Behav 22(4):269–289
Cherlin AJ, Kiernan KE, Chase-Lansdale L (1995) Parental divorce in childhood and demography outcomes in young adulthood. Demography 32(3):299–318
Cherlin AJ, Furstenberg FF, Chase-Lansdale L, Kiernan KE, Robins PK, Morrison DR, Teitler JO (1991) Longitudinal studies of effects of divorce on children in Great Britain and the United States. Science 252 (5011):1386–1389
Corak M (2001) Death and divorce: the long-term consequences of parental loss on adolescents. J Labor Econ 19(3):682–715
Cunha F, Heckman JJ, Lochner L, Masterov DV (2006) Interpreting the evidence on life cycle skill formation. In: Hanushek E, Welch F (eds) Handbook of the economics of education, vol 1. Elsevier Science, Amsterdam
Currie J, Tekin E (2006) Does child abuse cause crime? NBER Working Paper 12171, April, Cambridge MA, National Bureau of Economic Research
Dustmann C (2004) Parental background, secondary school track choice, and wages. Oxf Econ Pap 56(2):209–230
Elo IT, Preston SH (1996) Educational differentials in mortality: United States, 1979–1985. Soc Sci Med 42 (1):47–57
Ermisch JF, Francesconi M (2001) Family structure and children’s achievements. J Popul Econ 14(2):249–270
Ermisch JF, Francesconi M, Pevalin DJ (2004) Parental partnership and joblessness in childhood and their influence on young people’s outcomes. J R Stat Soc, A 167(1):69–101
Finlay K, Neumark M (2008) Is marriage always good for children? Evidence from families affected by incarceration. NBER Working Paper 13928, Cambridge MA, National Bureau of Economic Research
Francesconi M, Jenkins SP, Siedler T (2005) Childhood family structure and schooling outcomes: evidence for Germany. Working Paper 2005-22, Colchester, Institute for Social and Economic Research, University of Essex
Frick, J (2007) SOEP-Monitor. Zeitreihen zur Entwicklung von Indikatoren zu zentralen Lebensbereichen. DIW Berlin, Berlin
Gang IN, Zimmermann KF (2000) Is child like parent? Educational attainment and ethnic origin. J Hum Resour 35(3):550–569
Garfinkel I, Rainwater L, Smeeding TM (2005) Equal opportunities for children: social welfare expenditures in the english-speaking countries and western Europe. Focus 23(3):16–23
Gennetian L (2005) One or two parents? Half or step siblings? The effect of family structure on young children’s achievement. J Popul Econ 18(3):415–436
Ginther DK, Pollak RA (2004) Family structure and children’s educational outcomes: blended families, stylized facts, and descriptive regressions. Demography 41(4):671–696
Gruber J (2004) Is making divorce easier bad for children? The long-run implications of unilateral divorce. J Labor Econ 22(4):799–833
Haisken-DeNew JP, Büchel F, Wagner GG (1997) Assimilation and other determinants of school attainment in Germany: Do immigrant children perform as well as Germans? Vierteljahreshefte zur Wirtschaftsforschung 66(1):169–179
Hanushek EA (2002) Publicly provided education. In: Auerbach AJ, Feldstein M (eds) Handbook of public economics, vol 4. Elsevier Science, Amsterdam
Haveman R, Wolfe B (1995) The determinants of children’s attainments: a review of methods and findings. J Econ Lit 33(4):1829–1878
Heckman JJ, Vytlacil E (2004) Econometric evaluation of social programs. In: Heckman JJ, Leamer E (eds) Handbook of econometrics, vol 5. Elsevier Science, Amsterdam
Iacovou M (2002) Regional differences in the transition to adulthood. Ann Am Acad Polit Soc Sci 580(3):40–70
Jenkins SP, Schluter C (2002) The impact of family income during childhood on later-life attainment: evidence from Germany. Working paper 2002-20, Colchester, Institute for Social and Economic Research, University of Essex
Jeschek W (2000) General education and vocational training in east Germany and participation in education. Q J Econ Res 69(2):295–316
Jonsson JO, Gähler M (1997) Family dissolution, family reconstruction, and children’s educational careers: recent evidence for Sweden. Demography 34(2):277–293
Lang K, Zagorsky JL (2001) Does growing up with a parent absent really hurt? J Hum Resour 36(2):253–273
Lantz PM, House JS, Lepkowski JM, Williams DR, Mero RP, Chen J (1998) Socioeconomic factors, health behaviors, and mortality: results from a nationally representative prospective study of US adults. J Am Med Assoc 279(21):1703–1708
Levine PB, Zimmerman DJ (2005) Children’s welfare exposure and subsequent development. J Public Econ 89(1):31–56
Mahler P, Winkelmann R (2004) Single motherhood and (un)equal educational opportunities: evidence for Germany. IZA Discussion Paper 1391, Bonn, IZA
Manski CF (1990) Nonparametric bounds on treatment effects. Am Econ Rev Pap Proc 80(2):319–323
Manski CF, Sandefur GD, McLanahan S, Powers D (1992) Alternative estimates of the effect of family structure during adolescence on high school graduation. J Am Stat Assoc 87(417):25–37
Marmot MG, Shipley MJ, Rose G (1984) Inequalities in death—specific explanations of a general pattern? The Lancet 323 (8384):1003–1006, May
Martiny D, Schwab D (2002) Grounds for divorce and maintenance between former spouses: Germany. Available at: http://www2.law.uu.nl/priv/cefl/Reports/pdf/Germany02.pdf
Moffitt RA (2004) Introduction to the symposium on the econometrics of matching. Rev Econ Stat 86(1):1–3
Page ME, Stevens AH (2004) The economic consequences of absent parents. J Hum Resour 39(1):80–107
Painter G, Levine DI (2000) Family structure and youths’ outcomes: which correlations are causal? J Hum Resour 35(3):524–549
Piketty T (2003) The impact of divorce on school performance: evidence from France, 1968–2002. CEPR Discussion Paper 4146, December, London, Centre for Economic Policy Research
Pribesh S, Downey DB (1999) Why are residential and school moves associated with poor school performance? Demography 36(4):521–534
Riphahn RT (2005) Are there diverging time trends in the educational attainment of nationals and second generation immigrants? J Econ Stat (Jahrbücher für Nationalökonomie und Statistik) 225(3):325–346
Rosenbaum PR, Rubin DB (1985) Constructing a control group using multivariate matched sampling methods that incorporate the propensity score. Am Stat 39(1):33–38
Rosenzweig MR, Wolpin KI (1995) Sisters, siblings, and mothers: the effect of teen-age childbearing on birth outcomes in a dynamic family model. Econometrica 63(2):303–326
Ruhm CJ (2004) Parental employment and child cognitive development. J Hum Resour 39(1):155–192
Sanz-de-Galdeano A, Vuri D (2007) Parental divorce and students’ performance: evidence from longitudinal data. Oxf Bull Econ Stat 69(3):321–338
Sianesi B (2004) An evaluation of the Swedish system of active labor market programs in the 1990s. Rev Econ Stat 86(1):133–155
Siedler T (2006) Family and politics: does parental unemployment cause right-wing extremism? IZA Discussion Paper 2411, Bonn, IZA
Szydlik M (2000) Lebenslange Solidarität? Generationenbeziehungen zwischen erwachsenen Kindern und Eltern. Leske + Budrich, Opladen
Wagner M (1997) Scheidung in Ost- und Westdeutschland. Zum Verhältnis von Ehestabilität und Sozialstruktur seit den 30er Jahren. Campus Verlag, Frankfurt
Wojtkiewicz RA (1993) Simplicity and complexity in the effects of parental structure on high school graduation. Demography 30(4):701–717
Zentrum für Umfragen, Methoden und Analysen (ZUMA) (2005) System of social indicators for the Federal Republic of Germany: key indicators 1950–2005. Available at: http://www.gesis.org/EN/social_monitoring/social_indicators/Data/System/keyindic/ keyindicators.pdf
Financial support from the Anglo-German Foundation, the UK Economic and Social Research Council, DIW Berlin, and the University of Essex is gratefully acknowledged. We wish to thank the Editor (Christian Dustmann) and two anonymous referees for guidance and helpful suggestions. We are also grateful to Anders Björklund, Miles Corak, Monica Costa Diaz, John Ermisch, Steve Machin, Anna Vignoles, and seminar participants at HM Treasury (London), the Departments for Work and Pension and for Education and Skills (London), Institute of Education, Institute for Fiscal Studies, University of Essex, the 2006 Royal Economic Society and ESPE Conferences, and the CEPR Conference on Economics of Education and Education Policy in Europe (Uppsala) for helpful comments.
Responsible editor: Christian Dustmann
An erratum to this article can be found at http://dx.doi.org/10.1007/s00148-010-0327-7
About this article
Cite this article
Francesconi, M., Jenkins, S.P. & Siedler, T. Childhood family structure and schooling outcomes: evidence for Germany. J Popul Econ 23, 1073–1103 (2010). https://doi.org/10.1007/s00148-009-0242-y