Scarred for Life? Early-Life Experience of the Post-Reunification Economic Crisis in East Germany and Physical and Mental Health Outcomes in Early Adulthood Scarred for Life? Early-Life Experience of the Post-Reuni cation Economic Crisis in East Germany and Physical and Mental Health Outcomes in Early Adulthood Lara Bister, Jeroen Spijker, Fanny Janssen, Tobias Vogt Abstract: Existing research suggests adverse short-term health effects of economic crises during early life, yet, the long-term health effects for children and adolescents exposed to economic crises are still understudied. We investigated the early-adult health implications of experiencing the post-reuni cation economic crisis in East Germany in the early 1990s during infancy, childhood and adolescence. Using data from the German Socio-Economic Panel (SOEP) and its linkage with German pension records (SOEP-RV), we applied entropy balancing and conditional quantile regression to assess the relationship between the experience of the economic crisis during early life (ages 0-17) and physical and mental health effects (SF-12 summary scores) in N = 2,337 young adults (aged 17-30) from East and West Germany. Our results indicate mainly no signi cant physical health effects, yet, signi cant adverse mental health effects for respondents exposed to the economic crisis in East Germany, especially in young women of average and better mental health. Parental unemployment was an additional risk factor for young women’s mental health. Thus, we suggest women who experienced economic crises during early life are at increased risk of adverse mental health effects already early in their adulthood. Support for families in times of economic crisis and early prevention for infants, children and adolescents exposed to economic adversities is warranted. Keywords: Economic crisis · Life course approach · Family stress · SF-12 · Health · Germany 1 Introduction Economic stress is consistently associated with adverse physical and mental health outcomes, especially during economic crises (see Karanikolos et al. 2016; Mucci et al. 2016; and Thompson et al. 2019 for review). Children and adolescents are especially vulnerable to the adverse health effects of economic crises, as their Comparative Population Studies Vol. 48 (2023): 701-740 (Date of release: 14.12.2023) Federal Institute for Population Research 2023 URL: www.comparativepopulationstudies.de DOI: https://doi.org/10.12765/CPoS-2023-28 URN: urn:nbn:de:bib-cpos-2023-28en0 • Lara Bister, Jeroen Spijker, Fanny Janssen, Tobias Vogt702 health is sensitive to changes in their socio-economic environment (Elder 1998; Kuh et al. 2013) while being entirely dependent on their families and unable to mitigate the consequences of the emerging economic stress themselves (Elder 1999). This process may be explained by physiological dysregulation resulting from increased distress exposure (Glei et al. 2007) that may result in later-life adverse physical and mental health outcomes (Alastalo et al. 2013; Elder/Caspi 1988). Therefore, exploring the relationship between early-life exposure to economic crises and long- term health outcomes is particularly important for understanding the accumulation processes of poor health over the life course and explaining health inequalities in adulthood. Both empirical evidence and the theoretical view of the life course approach strongly suggest that exposure to a signi cant adverse life event – such as an economic crisis – early in life carries signi cant physical and mental health risks over the remaining life course. First, recent empirical studies conducted in high- income countries have found adverse short-term health effects among children and adolescents of both the general occurrence of an economic crisis (Anagnostopoulos/ Soumaki 2013; Cui/Zack 2013; Rajmil et al. 2013) and the family exposure to an economic crisis, i.e., through parental unemployment and family economic hardship (Briody 2021; Costa et al. 2020; Rajmil et al. 2013). In addition, studies on infant exposure to economic stress suggest higher mortality and cardiovascular health differences in later life compared to unexposed children (Alessie et al. 2019; van den Berg et al. 2006, 2011). Second, the theoretical assumptions of the life course approach imply mid- and long-term mental health effects of exposure to economic crisis during childhood and adolescence, considering the sensitive period model within the life course approach (Ben-Shlomo et al. 2014). Speci cally, children and adolescents are more susceptible to external stressors than adults; thus, early-life exposure can lead to long-term adverse health outcomes (Kuh et al. 2013). Economic stressors, in particular, increase family stress and ultimately harm children’s and adolescents’ physical and mental health (cf. Gard et al. 2020; Masarik/Conger 2017). In addition, recent studies on parental unemployment show that particularly school-age children and adolescents might suffer long-term mental health effects until adulthood (Moustgaard et al. 2018; Nikolova/Nikolaev 2018; Powdthavee/Vernoit 2013). While these results emphasise the vulnerability of children and adolescents to family economic stress, the adult health consequences of early-life economic adversities, particularly economic crises, remain unclear. This study seeks to contribute to this discussion by investigating the relationship between early-life exposure to an economic crisis, i.e., during infancy, childhood and adolescence, and physical and mental health in early adulthood. Speci cally, we study this relationship in the context of the post-reuni cation economic crisis in East Germany in the early 1990s (see 2.2. for more information). To our knowledge, research on the long-term health consequences of early-life exposure to the post- reuni cation economic crisis in East Germany is limited to one study on metabolic health effects (Bister et al. 2022). We are the  rst to study this economic crisis’ long- Scarred for Life • 703 term comprehensive mental and physical health consequences while considering not only the general but also the family-level exposure to economic stress. Accordingly, we address the following research questions: (1) What is the general early adult physical and mental effect of exposure to the post-reuni cation economic crisis in East Germany during infancy, childhood and adolescence? (2) What is the role of parental unemployment in this relationship, i.e., the direct exposure to the economic crisis? In assessing the exposure to this economic crisis in early life and subsequent physical and mental health consequences in early adulthood, we follow the life course approach by distinguishing between infancy, childhood and adolescence as the three main sensitive phases of the life course (Kuh et al. 2013). Furthermore, we conceptualise the adverse health effects of experiencing economic stress at the family level during economic crises for children and adolescents using the family stress model (see Masarik/Conger 2017 for review; Solantaus et al. 2004) and speci cally consider parental unemployment during the economic crisis period. Our results thereby contribute to the understanding of long-term health consequences of early-life economic adversities and further provide insights into the setting of the post-reuni cation economic crisis in East Germany, which we utilise for our study and which has been widely understudied. 2 Background 2.1 Theoretical considerations In our efforts to investigate the early-adult health implications of experiencing severe economic stress during early life, we rely on two theoretical concepts, namely the family stress model and the life course approach. First, the Family Stress Model (FSM) broadly explores how external stressors affect family dynamics (Masarik/Conger 2017) focusing on traditional family constellations, i.e. a two-generation household consisting of one or two parent(s) and at least one of their dependent children. The mechanism of action is described as follows: the external stressor is initially absorbed by the parents. In processing the related adversities of this stressor, they experience increased distress leading to disruptions in the family dynamics and eventually to a decline in parenting. This decline in parenting entails the signi cantly decreasing ability to ful l parental duties to varying extents, ranging from lacking emotional support to child neglect (Masarik/Conger 2017). Subsequently, the family stress transmitted from the parents to their children may negatively impact child development, potentially resulting in adverse health consequences such as developmental delays or immense physical and mental health problems (cf. Gard et al. 2020; Solantaus et al. 2004). Economic stressors, such as poverty, unemployment, or  nancial strain, are amongst the common stressors studied when it comes to applications of the FSM in empirical research. Speci cally, economic stressors at the family level triggered • Lara Bister, Jeroen Spijker, Fanny Janssen, Tobias Vogt704 by worsened economic conditions – i.e., low income, debts, or other adverse  nancial events – may lead to decreased parenting capabilities in the parents (cf. Schmiedeberg/Bozoyan 2021) and subsequently increased psychological distress in children (Conger et al. 2012; Conger/Conger 2002). In turn, children are at higher risk of suffering from adjustment problems (see Masarik/Conger 2017 for review), leading to developmental delays and subsequently adverse long-term health outcomes. While the FSM may aid in conceptualising how economic stressors during an economic crisis, it does not explain the relationship between early-life economic adversities and later life health consequences. Therefore, we additionally utilise the life course approach. The life course approach (Brandt et al. 2012; Kuh et al. 2013) stresses the signi cance of early-life conditions for the healthy development of individuals until and throughout adulthood. Speci cally, it emphasises children’s and adolescents’ emotional, social, and physical dependence on their social environment, such as their family. Exposure to sources of family economic stress during the vulnerable early years of life, such as an economic crisis during infancy, childhood or adolescence, may accumulate adverse health outcomes over the life course, into signi cant health differences in adulthood (Kuh et al. 2013). Hence, the experience of an economic crisis and its subsequent family economic stress during early life may play a critical role in individuals’ long-term health. 2.2 Effects of economic crises on child and adolescent health Previous research has shown that economic crises are particularly severe for children and adolescents‘ short-term physical and mental health (see Rajmil et al. 2014 for review). For mental health, Costa and colleagues (2020) studied children aged 7 to 10 born shortly before or during the Great Recession of 2008 in Portugal. They found that children directly exposed to the economic crisis showed increased psychosocial functioning problems, poorer health-related quality of life, and increased symptoms of depression, anxiety, and stress. Similarly, Rajmil and colleagues (2013) found in Spain that parental unemployment during this economic crisis immediately and adversely affected the health-related quality of life of children aged 0-12. In their research on US adolescents, Cui and Zack (2013) observed that mental health declined during the Great Recession. Finally, Anagnostopoulos and Soumaki (2013) showed increased uptake of psychiatric services in children and adolescents in Greece during the same economic crisis. At the same time, research on the physical health effects of economic crises for children and adolescents is scarce. Of the previously discussed studies, Cui and Zack (2013) are the only ones that also examined physical health outcomes and showed declined self-rated health but no signi cant decreases in physical health in adolescents. Rajmil and colleagues (2013) also studied children’s body mass index and found an increased prevalence in child adiposity after exposure to the economic crisis. These  ndings are consistent with studies investigating children’s weight gain after the Great Recession of 2008 in Portugal (Rodrigues et al. 2021) and Ireland (Briody 2021). Both concluded a positive association between Scarred for Life • 705 deteriorated nutritional status and decreasing physical activity in families affected by an economic crisis and children’s weight gain. The fact that children and young people are affected by the onset of economic crises is generally unsurprising. Several studies suggest that the sole onset of an economic crisis can lead to psychological distress in the population (see Mucci et al. 2016 for review). This general negative health effect likely evolves from higher levels of economic insecurity during economic crises that, in turn, reinforce economic stress emerging from any occurring unemployment (Hessel et al. 2014; Reibling et al. 2017). Parents, in particular, may experience increased distress as they experience a greater sense of responsibility for their children and thus might feel strongly threatened by an economic crisis. Accordingly, a study by Chzhen (2016) showed that the presence of children in the household increased the probability of reporting distress in the context of the Great Recession of 2008 in Europe. Consequently, and based on the Family Stress Model (FSM), parental distress experiences during economic crises may have a negative impact on their children. Solantaus and colleagues (2004) validated the assumptions of the FSM for children’s mental health during the 1990s economic crisis in Finland. Their study showed that increased economic pressure and negative changes in parental mental health, marital interaction, and parenting quality contributed to children’s mental health risks. At the same time, a family’s socio-economic capability is crucial for the coping with the adverse economic conditions, and may decide on adverse health effects in the children (Currie/Stabile 2002; Doyle et al. 2007). Speci cally, parental educational attainment (Rajmil et al. 2013) or household income (Burgess et al. 2004) might facilitate physical and mental child health inequalities induced by family economic stress. Accordingly, we hypothesise that children from families that were more severely affected by economic stress show worse physical and mental health in early adulthood. 2.3 Adult health effects of economic adversities during early life Studies on adult health effects of exposure to economic crises in early life are scarce. Only one study by Bister and colleagues (2022) investigates early-adult metabolic health effects of early-life exposure to an economic crisis. Speci cally, they studied several metabolic health outcomes of East Germans born shortly before or during the post-reuni cation economic crisis. They found increased early-adult metabolic health risks in women exposed to the economic crisis in utero and infancy. However, they only analysed the overall burden of the economic crisis and did not consider experiences of family economic stress through, e.g., parental unemployment. Following the insights of the life course approach, we hypothesise that exposure to an economic crisis during early life increases the long-term physical and mental health risks. While research on the long-term health consequences of economic crises during early life is scarce, evidence from studies of parental unemployment provides support. Several studies showed adverse long-term physical and mental health • Lara Bister, Jeroen Spijker, Fanny Janssen, Tobias Vogt706 outcomes after experiencing economic hardship and parental unemployment at a young age (Lindström et al. 2012, 2014; Moustgaard et al. 2018; Nikolova/ Nikolaev 2021; Widding-Havneraas/Pedersen 2020). For example, Moustgaard and colleagues (2018) observed increased psychotropic medication use in adolescents experiencing paternal unemployment in Finland with long-lasting effects. Nikolova and Nikolaev (2021) found decreases in life satisfaction during early adulthood (ages 18-31) for individuals experiencing parental unemployment during childhood. Lindström and colleagues (2012, 2014) con rmed in their research in Sweden that childhood and adolescent economic stress (ages 0-18) accumulates to poor self- rated general health (Lindström et al. 2012) and mental health (Lindström et al. 2014) in adulthood. Widding-Havneraas and Pedersen (2020) con rmed these  ndings for 20 European countries and found an association between childhood and adolescent economic stress (ages 0-18) and adult self-rated health and cancer. Accordingly, we hypothesise that the exposure to severe early-life economic stress is associated with worse physical and mental health in early adulthood. 2.4 The post-reuni cation economic crisis in East Germany and its health effects In this paper, we study the relationship between exposure to an economic crisis during childhood and adolescence and health in early adulthood using the post- reuni cation economic crisis in East Germany in the early 1990s. This economic crisis followed the sudden economic restructuring in East Germany following the German reuni cation in 1990, during which the previously centrally planned East German economy was transformed into a competitive market economy (Burda/ Hunt 2001). This process caused many East German companies to go bankrupt, resulting in mass unemployment in entire East Germany, where full employment had been a political imperative before the reuni cation (Burda/Hunt 2001; Chevalier/ Marie 2017). This signi cant economic shock in East Germany occurred within a few months and suddenly exposed more than 16 million East Germans to a severe economic crisis that escalated between 1990 and 1994 (Federal Statistical Of ce 2020a). Although West Germany also saw a slight increase in unemployment, the changes were not comparable to the substantially rising unemployment rate in East Germany in combination with the overall decline of the East German economy. Figure 1 illustrates the tremendous economic growth and unemployment rate differences between East and West Germany. We identi ed the main economic crisis period between 1990 and 1994 with a sudden drop in the gross domestic product (GDP) per capita in 1990 that only recovered to pre-reuni cation levels in 1995 (Fig. 1, left panel), and a sharp increase in unemployment from 1990 that slightly decreased again in 1995 (Fig. 1, right panel). The variation in the economic situation between East and West Germany was particularly evident in the unemployment statistics: in 1993, for example, only about 7 percent of the West German population was directly or indirectly affected by unemployment compared to approximately 27 percent of the East German Scarred for Life • 707 population (Federal Statistical Of ce 2020b;Hauser 1996; Hofer et al. 1995). The rapid increase in unemployment rates in East Germany was not only a macroeconomic shock but also a completely new experience for East Germans who were not used to unemployment prior to the German reuni cation, neither its anticipation nor its consequences. Different from the GDP, however, the unemployment rates did not recover after 1994 but established themselves and introduced a decade of futile labour market prospects in East Germany. Although the post-reuni cation economic crisis in East Germany was a signi cant economic shock for the entire East German population, its health effects have received relatively little attention, especially those of children and adolescents exposed to it. Previous research mainly focused on East-West German mortality differences and reported general East German health disadvantages in the older Fig. 1: Comparison of the GDP per capita and the unemployment rate during pre-reuni cation period (1960-1989), pre-reuni cation economic crisis (approx. 1990-1994, in grey) and later years between East and West Germany 0 5 10 15 20 25 1980 1985 1990 1995 2000 year East Germany West Germany Children present in household in East Germany* Children present in household in West Germany 100 125 150 175 200 225 250 1960 1970 1980 1990 2000 year 1990 US$ (PPP-Geary Khamis) GDP per capita: relative levels (1960=100) Unemployment rate (SOEP) Percentage (%) unemployed Note: Gross Domestic Product (GDP) per capita (1960-1997) in 1990 US$ converted at Geary Khamis purchasing power parity (GK PPP) and compared to 1960 (1960 = 100) (left graph); unemployment rate (1984-1999) for total population and for households, in which children below the age of 16 are present, for East and West Germany. *No unemployment data available for pre-reuni cation East Germany (1983-1989); however, since full employment was a political imperative in pre-reuni cation socialist East Germany (cf. Burda/Hunt 2001), we assume the unemployment rate from 1983 to 1989 being zero; the grey area indicates years of post-reuni cation economic crisis in East Germany from 1990-1994. Source: GDP per capita data from The Conference Board Total Economy DatabaseTM (2013); unemployment rates (own calculations) from the German Socio-Economic Panel (SOEP), Liebig et al. (2019). • Lara Bister, Jeroen Spijker, Fanny Janssen, Tobias Vogt708 population (Grigoriev/Pechholdová 2017; Nolte/McKee 2004; Vogt 2013). Some of these health differences can be attributed to the inherently different public health system in pre-reuni cation East Germany (Busse/Riesberg 2005; Lampert et al. 2019). However, the role of the post-reuni cation economic shock in East Germany remains unknown. The prolonged economic insecurity in East Germany after the reuni cation and its in uence on East Germans’ health has been frequently studied (Berth et al. 2011; Easterlin/Plagnol 2008; Frijters et al. 2005). However, these studies mainly focused on short-term health effects for the adult East German population, and the younger population remained largely unresearched. Only two studies addressed the health of younger East Germans. First, Forkel and Silbereisen (2001) researched East German adolescents’ well-being in 1993, i.e., during the post-reuni cation economic crisis, concerning parental unemployment, parental mood, and family climate. Their results did not indicate signi cant health disadvantages for East German compared to West German adolescents. However, they adopted a cross-sectional study design, which only provided a snapshot during one year of the post-reuni cation economic crisis and did not account for individual or family-level confounding or a longitudinal perspective. Second, Bister and colleagues (2022) observed in their recent study increased long-term metabolic health risks of the post-reuni cation economic crisis. However, they did not consider parental unemployment, i.e., the direct experience of the economic crisis. Thus, the full extent of the long-term mental and physical health consequences of this economic crisis for children and adolescents, particularly concerning the experience of family economic stress, remains unclear. 3 Methods 3.1 Data and sample We used data from the German Socio-Economic Panel (SOEP) (Liebig et al. 2019), a nationally representative longitudinal study of private households in Germany. The survey, which includes more than 25,000 respondents in nearly 15,000 households, has been conducted annually since 1984 in West Germany and 1990 in East and West Germany. With an East German sample compiled already in June 1990, the SOEP is the only longitudinal study enabling the study of East German families during the post-reuni cation period introducing disrupting political and economic transformation to an entire society (Liebig et al. 2019). The SOEP core study contains detailed personal, social, and economic information for all household members aged 16 and older (Wagner et al. 2007). To identify the pre- and post-reuni cation household residence in East or West Germany, we additionally retrieved information from the SOEP regional data (Knies/Spiess 2007) (identi cation of 84.5 percent of the  nal sample) and the SOEP-RV data linkage with longitudinal data on social security pension records from the German Pension Insurance (“Deutsche Rentenversicherung”) (Lüthen et al. 2022) (identi cation of 16.5 percent of the  nal Scarred for Life • 709 sample). These additional data sets complement the comprehensive survey data of the SOEP with objective information from survey metadata (SOEP regional data) and administrative registers (SOEP-RV) and are thus highly reliable. To summarise, for the purpose of our study we utilised the SOEP survey year of 1990 and the pre-1990 information from the SOEP-RV data link to identify the place of residence at the onset of the post-reuni cation economic crisis in East Germany (with West Germany as reference); the 1990-1994 SOEP survey years and respective SOEP-RV information to reconstruct the family-level exposure during the economic crisis years; and the 2002-2009 SOEP survey years to retrieve the SF-12 variables informing on physical and mental health. Thereby, the SOEP enabled us to study the long-term physical and mental health of a vastly understudied population, i.e., East German cohorts experiencing the post-reuni cation economic crisis as children or adolescents. The SOEP thereby aided in depicting the long-term health consequences of exposure to extreme economic stress for these cohorts and their families. Subsequently, our premise in the sample selection was to obtain as large a sample as possible with high statistical power to address our objective. We restricted our sample to respondents born between 1973 and 1989, i.e., those who were between zero and 17 years old at the onset of the post-reuni cation economic crisis in 1990. We refrained from including individual respondents born after the onset of the post-reuni cation economic crisis to avoid bias due to birth selection in post-reuni cation East Germany (cf. Adler 1997; Huinink/Kreyenfeld 2006). We only considered respondents with,  rst, at least one parent included as a respondent themselves in the SOEP and, second, the parental household residence in either East or West Germany before the German reuni cation. Since neither the SOEP regional data nor the SOEP-RV differentiates between East and West Berlin, we excluded Berlin. Subsequently, our  nal sample consisted of N = 2,337 individuals from East (n = 982, 42.0 percent) and West Germany (n = 1,355, 58.0 percent) (see Table 1 for sample description and Figure S1 for sample selection). 3.2 Health outcome variables We measured two dependent variables indicating the physical and mental health of the respondents using the SF-12 summary scores for physical and mental health functioning (SF-12) (SOEP-version, see Andersen et al. 2007 and Appendix page 37). The SF-12 physical (PCS) and mental component scores (MCS) were computed from eight domains: namely, physical role limitation, general health perception, vitality, bodily pain, physical functioning, social function, socio-emotional role limitations, and mental health (RAND Corporation 2020). The SF-12 is known as a valid and reliable tool for assessing physical and mental health based on subjective health measures (Andersen et al. 2007; RAND Corporation 2020) and has previously been used to measure health functioning in adolescents aged 13-19 and young adults aged 20-23 (cf. Ge et al. 2019; Jörngården et al. 2006; Moscarelli/Manning 2009) on a scale from zero (poor health) to 100 (good health) (Turner-Bowker/Hogue 2014). We included both the PCS and the MCS as continuous variables observed once per respondent at the earliest available age in young adulthood to avoid unobserved • Lara Bister, Jeroen Spijker, Fanny Janssen, Tobias Vogt710 heterogeneity due to the accumulation of additional health risks over their adult life. Speci cally, we measured the respondents’ health outcomes between ages 17 (the minimum response age for the health measure) and 30 (the earliest available response age for the 1973 cohort in 2002). 3.3 Main explanatory variables We measured the experience of the post-reuni cation economic crisis in East Germany using two variables, namely the region of the pre-reuni cation residence and the experience of migration to West Germany. For the  rst measure, we created Tab. 1: Sample description of the unbalanced sample (i.e., with no balancing weights from entropy balancing applied) of the full N = 2,337 sample Variable Mean / % Std. dev. Min Max Health outcomes during early adulthood (using SF-12) Mental Component Score (MCS)+ 49.56 9.56 13.17 71.81 Physical Component Score (PCS)+ 56.18 6.49 13.63 73.61 Age at SF-12 in years 21.00 3.60 17 30 Parental household residence before the German reuni cation (%) West Germans (n = 1,355) 57.98% East Germans (n = 982) 42.02% migrated to West Germany* 29.74% stayed in East Germany* 70.26% Parental unemployment 1990-94 (%) Total 31.66% West German (n = 1,355) 20.52% East German (n = 982) 47.05% Demographics Birth year 1982.23 4.51 1973 1989 Female (%) 50.02% Age at onset of the economic crisis (1990) 7.7 7 4.51 0 17 in age groups in 1990 (%) Infancy (aged 0-2) 21.31% Early childhood (aged 3-5) 22.12% Childhood (aged 6-11) 38.30% Adolescence (aged 12-17) 18.27% Number of siblings 1.49 1.26 0 9 Mother’s age at birth in years 25.97 5.05 12 48 Same residence region at health outcome 91.62% + SF-12 Physical Component Score (PCS) and Mental Component Score (MCS) are scale measures that may potentially range from 0 (poor health) to 100 (very good health). * Percentage of East German subsample migrating to West Germany or staying in East Germany during the 1990s. Source: German Socio-Economic Panel (SOEP) and SOEP-RV link Scarred for Life • 711 a binary variable indicating the respondents’ parental pre-reuni cation residence in East Germany (reference: West Germany) identi ed by either their household residence retrieved from the SOEP regional data or by their employment location retrieved from the SOEP-RV data (see Data and sample). For the second measure of migration from East to West Germany, we created a binary variable indicating the migration from the pre-reuni cation residence in East or West Germany to the opposite region (reference: no migration) between 1990 and 1994. With this measure, we aimed to consider the large share of East Germans migrating to West Germany during the early 1990s (Heiland 2004), which undeniably shortened the exposure period to the post-reuni cation economic crisis in East Germany. We included this measure as an interaction with the exposure to the post-reuni cation economic crisis to account for the sole exposure to the initial post-reuni cation economic shock (i.e., the initial region of residence in East Germany and subsequent migration to West Germany (Liebig et al. 2019) and shortened exposure to an economic crisis), or full exposure to the severe economic crisis in post-reuni cation East Germany (i.e., no migration). We further assessed the respondents’ family’s direct exposure to the economic crisis by including a binary variable on their experience of any spell of parental unemployment during the economic crisis period between 1990 and 1994 (reference: no parental unemployment). We base this information on paternal and maternal unemployment spells between January 1990 and December 1994. We retrieved these from parental employment biographies in either the SOEP core study for households that entered the SOEP in 1990 or earlier or the social security pension records (SOEP-RV) for households that entered the SOEP later than 1990. We interacted the parental-unemployment variable with the above-mentioned binary variable indicating the respondent’s parental pre-reuni cation residence to account for different meanings in parental unemployment during the post-reuni cation economic crisis period between East and West Germany. 3.4 Control variables As control variables, we included  rst, a categorical measure for the exposure age at the onset of the post-reuni cation economic crisis in 1990 with categories of “Infancy” (age 0-2, reference), “Early childhood” (age 3-5), “Childhood” (age 6-11), and “Adolescence” (age 12-17); second, a continuous measure for the number of siblings; third, the respective other SF-12 PCS or MCS value to control for the reciprocal in uence of physical and mental health; and, fourth, a continuous measure for the mothers’ age at  rst childbirth, z-standardised for East and West Germany respectively. The  nal control variable served as proxy for the respondents’ socio-economic background. More commonly used measures such as the parental education comparably showed many missing values for our sample (see Table A1) or relied on respondent-reported approximations (see Fig. A2). To maintain a larger sample and achieve higher statistical power, we therefore decided to measure the socio- economic background using the mothers’ age at  rst childbirth. This measure has • Lara Bister, Jeroen Spijker, Fanny Janssen, Tobias Vogt712 been used before as a proxy for socio-economic background (van Roode et al. 2017) and general effects on later-life health outcomes (Carslake et al. 2017). We, however, tested the robustness of our variable selection in a sensitivity analysis replacing mothers’ age at  rst childbirth with the highest parental education leading to nearly congruent results and thus con rming our modelling choice (see Fig. A2). Moreover, the included control variable of the exposure age at the onset of the post-reuni cation economic crisis in 1990 additionally captures the respondents’ ages at the health outcome assessment and thereby partially accounts for the monitoring gap between the exposure and the health outcome assessment. Corresponding to the SF-12, the survey questions were roughly collected at the same time for the vast majority of our sample, i.e., in the years of 2002 (n = 1,549, 66.3 percent), 2004 (n = 427, 18.3 percent), and the years thereafter (n = 361, 15.4 percent). With a collinearity coef cient of 0.7847 (p<0.0001), the exposure age to the economic crisis should thereby re ect on the age at the respondents’ SF-12 assessment. 3.5 Analytical approach In our analysis, we proceeded in two steps (Table 2). In Step 1, we applied entropy balancing to adjust the covariate distribution of the control group data (i.e., the West German sample) by re-weighting the observations to improve the comparability of both groups regarding the covariate distribution in the treatment group (i.e., the East German sample). Based on known sampling moments, entropy balancing improves the covariate balance and helps to adjust survey samples to the known characteristics of a target population (Hainmueller 2012). We calculated the entropy balancing sample weights in StataSE 17 using “ebalance” (Hainmueller/Xu 2013). We matched the East and the West German respondents on (1) their parental unemployment experiences, (2) their birth year, (3) their gender, (4) their age at the SF-12 survey, (5) their number of siblings, and (6) their mothers’ maternal age at childbirth (z-standardised for East and West German samples separately) (see Table A1 for matching statistics). Using the obtained sampling weights, we re-weighted the observations for the analysis in Step 2. With this re-weighting strategy, we refrained from achieving results representative of the East and West German population structure. However, we aimed to obtain two comparable samples (in at least the speci c attributes used) and enable the isolation of the exposure effect of the post-reuni cation economic crisis. In Step 2 of our analysis, we applied conditional quantile regression with a re-weighted sample using the entropy balancing weights from Step 1 (Table 2). Conditional quantile regression is an extension of linear regression and estimates the conditional mean within a quantile of a response variable across the values of a predictor variable. Speci cally, we used conditional quantile regression to identify the heterogeneous effects of the experience of the post-reuni cation economic crisis on physical (PCS) and mental health (MCS) for the quantiles of their respective distribution. We decided for quantile regression as our main analytical strategy, Scarred for Life • 713 as it allows to estimate a linear relationship while accounting for different slopes depending on the quantile in the outcome variable distribution. It thereby allows for a more ef cient inclusion of data outliers compared to mean-based regression techniques such as ordinary least squares regression (cf. Rios-Avila/Maroto 2022). Speci cally, we suggested that the effect of the experience of the post-reuni cation economic crisis in East Germany during childhood and adolescence on the young adult physical (PCS) and mental health outcomes (MCS) differed over different quantiles of the conditional distribution. Therefore, we  t quantile regression at the 20th, 40th, 60th, and 80th percentiles of the conditional distribution of the exposure to the economic crisis in East Germany on both health outcomes PCS and MCS. 4 Results 4.1 Descriptive results We observed in two-sample t-tests that East Germans scored on average 0.3 units lower in the PCS (p=0.2440) and 2.2 units signi cantly lower in the MCS (p<0.001) compared to the West German sample. The descriptive analysis further Tab. 2: Overview of the analytical strategy using entropy balancing (Step 1) and conditional quantile regression (Step 2) Used measures Step 1: Entropy balancing Step 2: Conditional quantile regression models Aim To balance the sample to To obtain the predicted average improve the comparability of change in SF-12 measures for the East and the West quantiles of the PCS/MCS German subsamples. distribution applying sample weights from Step 1. SF-12 PCS [No conditioning on SF-12 Health outcome variables SF-12 MCS measures in Step 1.] Pre-reuni cation residence (X) “Treatment” measure Internal East-West migration (Main) explanatory variables Parental unemployment Variables used for matching  predictors conditioning the treatment probability Age at economic crisis Gender Control variables Age at SF-12 assessment Maternal age at  rst childbirth N 2,337 2,337 Note: SF-12 summary score for health functioning shown as physical (SF-12 PCS) and mental component scores (SF-12 MCS); other measures correspond to descriptions in Health outcome variable and Main explanatory variables. Source: Own design • Lara Bister, Jeroen Spijker, Fanny Janssen, Tobias Vogt714 showed that 47.0 percent of the East German respondents experienced parental unemployment between 1990 and 1994, which was signi cantly higher than the West German respondents with 20.5 percent (p<0.001). The health scores for those who experienced parental unemployment between 1990 and 1994 were signi cantly lower with 1.19 units for the PCS (p<0.001) and 0.95 units for the MCS (p<0.5) for the total sample. However, we only found signi cantly lower health scores for East Germans who experienced parental unemployment during that time for the PCS (1.68, p<0.001), yet not for the MCS (0.33, p=0.6015). Table 3 shows the differences in early adult physical and mental health of the East German sample compared with the West German sample over the 20 percent, 40 percent, 60 percent and 80 percent quartiles of the distribution using mean difference testing (t-test results). Again, we observed no signi cant difference between the PCS of the East and West German samples across all quantiles, with lower PCS only in the 60 percent and 80 percent quantiles. The MCS, however, were smaller for the East German sample across all quantiles, with signi cant differences at the 5 percent level for the 60 percent and 80 percent quantiles, indicating health differences especially among those respondents at the upper end of the distribution of mental health scores. To fully understand how the experience of the post-reuni cation economic crisis in East Germany during early life shaped individuals’ physical and mental health in early adulthood, we turned to the multivariate analysis. Tab. 3: Differences in early adult physical and mental health of the East German sample compared with the West German sample over the 20 percent, 40 percent, 60 percent and 80 percent quartiles of the PCS/ MCS distribution Quantile 20% 40% 60% 80% Means per subsample Physical Component Score (PCS) West German 45.51 50.23 52.70 54.63 East German 46.57 50.33 52.58 54.30 p-value (T-test) 0.0547 0.8084 0.7001 0.2350 Mental Component Score (MCS) West German 35.08 40.50 44.24 47.28 East German 34.33 39.65 43.30 45.88 p-value (T-test) 0.1525 0.0548 0.0217 0.0003 Note: Mean differences (t-test results) of the health outcome variables SF-12 Physical Component Score (PCS) and Mental Component Score (MCS) over 20 percent, 40 percent, 60 percent and 80 percent quartiles of the distribution for the full N = 2,337 sample (no weighting applied). Annotations: italic font indicates signi cant differences (p<0.05) between West German and East German quantile-speci c mean. Source: German Socio-Economic Panel (SOEP) and SOEP-RV link Scarred for Life • 715 4.2 Regression results Figures 2 and 3 present the association between the experience of the post- reuni cation economic crisis during early life and early adult physical (PCS) and mental (MCS) health outcomes, along with the distribution of PCS and MCS (as average marginal effects, complete regression and effect tables available in Tables A3). In doing so, we structured our main results around three aspects, for physical and mental health outcomes respectively:  rst, the general experience of the onset of the economic crisis in 1990 (indicated by the birthplace in East or West Germany prior to the German reuni cation in 1990); second, the exposure to the post-reuni cation economic crisis throughout its duration from 1990 to 1994 (indicated by an interaction of the birthplace and an East-West migration during the economic crisis years); and third, the experience of parental unemployment between 1990 and 1994. First, we present our results on the general experience of the onset of the economic crisis in 1990 (research question 1). The quantile regression showed lower PCS for East German men across all quantiles ranging from -1.2 in the 20 percent quantile to -0.3 in the 60 percent quantile lower PCS compared to West German men. Nevertheless, none of the effects were statistically signi cantly different from zero (p>0.05). We observed almost no difference in PCS for East German women compared to West German women (Fig. 2). For MCS (Fig. 3), we found consistently lower scores for East German men and women than their West German counterparts. MCS differences for men ranged from -0.8 in the 80 percent quantile to -2.6 in the 60 percent quantile lower for East Germans, with signi cant effects at the 5 percent level only for the 60 percent quantile (p<0.01). For women, the MCS were from -2.0 in the 20 percent quantile to -3.4 in the 40 percent quantile lower for the East German samples, with statistically signi cant effects at the 5 percent level for the 40 percent (p<0.01), 60 percent and 80 percent quantiles (p<0.001). Second, we show our  ndings on the exposure to the post-reuni cation economic crisis, which showed increased East-West migration (research questions 1). The quantile regression results for the region of birth in East Germany that interacted with migration to West Germany showed no signi cant differences for PCS in men and women who migrated and did not migrate compared to their West German counterparts (Fig. 2). For MCS (Fig. 3), we observed signi cant effects at the 5 percent level for East German men only for those exposed to the economic crisis throughout its duration (i.e., no migration), with effects ranging from -1.5 in the 80 percent quantile to -3.1 in the 60 percent quantile decreased MCS, all statistically signi cant (p<0.05/p<0.01) except for the 20 percent quantile. For East German women, we found signi cant effects at the 5 percent level on the MCS for those who did and did not migrate to West Germany with similar effects. Third, we assessed the experience of parental unemployment during the economic crisis years between 1990 and 1994. The quantile regression results for the region of birth in East Germany that interacted with parental unemployment experience in the respective period revealed no statistically signi cant effects at the 5 percent level for PCS (Fig. 2), neither for East German men nor women. However, • Lara Bister, Jeroen Spijker, Fanny Janssen, Tobias Vogt716 the PCS of East German men experiencing parental unemployment was slightly lower than those who did not experience parental unemployment. The effects on the PCS in women tended towards zero. For MCS (Fig. 3), we found an overall trend of lower scores for both East German men with and without experience of parental unemployment during the economic crisis, with most scores not statistically signi cant. Similarly, the MCS for women were also overall lower for those who did and did not experience parental unemployment compared to their West German counterparts. However, the effects for those who experienced parental Fig. 2: Overall effect of the exposure to the post-reuni cation economic crisis in East Germany on physical health scores (SF-12 PCS) over four quantiles (20 percent, 40 percent, 60 percent and 80 percent) of the PCS distribution, estimated from conditional quantile regression models 20 40 60 80 Overall effect: exposure to the economic crisis East Germany (ref.: West Germany)in Average difference in PCS Average difference in PCS Overall effect X migration to West Germany (ref.: West Germany) -6 -4 -2 0 2 20 40 60 80 percent quantiles No migration Migration Average difference in PCS Overall effect X parental unemployment (ref.: West Germany) -6 -4 -2 0 2 20 40 60 80 20 40 60 80 No unemployment Unemployment percent quantiles percent quantiles Average difference in PCS Average difference in PCS -4 -2 0 2 4 20 40 60 80 percent quantiles -4 -2 0 2 4 20 40 60 80 20 40 60 80 No migration Migration Average difference in PCS -4 -2 0 2 4 20 40 60 80 20 40 60 80 No unemployment Unemployment percent quantiles percent quantiles Women Men -6 -4 -2 0 2 20 40 60 80 Note: Reference for each effect is pre-reuni cation residence in West Germany. Effects were calculated based on quantile regression models adjusted for East-West Germany migration and parental unemployment (both interacted with the pre-reuni cation residence), age at onset of the post-reuni cation economic crisis, children’s gender, their age at SF-12 assessment, and their mother’s age at  rst childbirth. See Tables A3 for full regression and effect tables. Source: German Socio-Economic Panel (SOEP) and SOEP-RV link Scarred for Life • 717 unemployment were far larger compared to those who did not, ranging from -1.2 in the 20 percent quantile to -5.2 in the 60 percent quantile, with statistically signi cant effects at the 5 percent level for all (p<0.05/p<0.01) except the 20 percent quantile. Fig. 3: Overall effect of the exposure to the post-reuni cation economic crisis in East Germany on mental health scores (SF-12 MCS) over four quantiles (20 percent, 40 percent, 60 percent and 80 percent) of the MCS distribution, estimated from conditional quantile regression models Note: Reference for each effect is pre-reuni cation residence in West Germany. Effects were calculated based on quantile regression models adjusted for East-West Germany migration and parental unemployment (both interacted with the pre-reuni cation residence), age at onset of the post-reuni cation economic crisis, children’s gender, their age at SF-12 assessment, and their mother’s age at  rst childbirth. See Tables A3 for full regression and effect tables. Source: German Socio-Economic Panel (SOEP) and SOEP-RV link 20 40 60 80 Overall effect: exposure to the economic crisis in East Germany (ref.: West Germany) Average difference in MCS Average difference in MCS Overall effect X migration to West Germany ( )ref.: West Germany -10 -5 0 5 20 40 60 80 percent quantiles No migration Migration Average difference in MCS Overall effect X parental unemployment (ref.: )West Germany -10 -5 0 5 20 40 60 80 20 40 60 80 No unemployment Unemployment percent quantiles percent quantiles Average difference in MCS Average difference in MCS -10 -5 0 5 20 40 60 80 percent quantiles No migration Migration Average difference in MCS -10 -5 0 5 20 40 60 80 20 40 60 80 No unemployment Unemployment percent quantiles percent quantiles Women Men -10 -5 0 5 20 40 60 80 -10 -5 0 5 20 40 60 80 20 40 60 80 • Lara Bister, Jeroen Spijker, Fanny Janssen, Tobias Vogt718 5 Discussion 5.1 Summary of results This study analysed the relationship between exposure to the economic crisis in post-reuni cation East Germany during infancy, childhood and adolescence (ages 0-17) and physical and mental health in early adulthood (ages 17-29). We sought to investigate (1) the general physical and mental health responses to the experience of an economic crisis while exploring the in uence of the initial economic shock immediately after the German reuni cation in 1990 and the following economic crisis in the early 1990s, and (2) the in uence of parental unemployment during this economic crisis, i.e., the direct exposure to this economic crisis. Our results indicate no signi cant early-adult physical health effects of the post- reuni cation economic crisis for East German-born men and women compared to their West German counterparts. However, we observed signi cantly worse mental health in East German men and women, especially among women of average and better mental health. Furthermore, when differentiating between the experience of the initial economic shock of the post-reuni cation economic crisis and the experience of the subsequent economic crisis from 1990 to 1994 (using an indicator of migration to West Germany), we only found signi cantly lower mental health for East German men who experienced the economic crisis throughout its duration. We further found signi cantly lower mental health for East German women who experienced parental unemployment during the post-reuni cation economic crisis compared to their West German counterparts, especially for those of average and better mental health. 5.2 Discussion of results Our  ndings emphasise the adverse long-term mental health effects of early- life economic adversities. In line with previous research, this result con rms our hypothesised negative relationship between early-life exposure to the economic crisis and early-adult health, as derived from previous research on immediate health and stress responses of children and adolescents to economic crises (Anagnostopoulos/Soumaki 2013; Cui/Zack 2013; Rajmil et al. 2013). Furthermore, our study provides evidence that women in particular carry the long-term mental health consequences after experiencing an economic crisis early in life, both in general and through parental unemployment at the family level. These  ndings align with prior studies showing stronger reactions to external stressors in women than in men and increased hormonal and physical dysregulation (Dedovic et al. 2009), which may also affect their mental health. Similarly, Bister and colleagues (2022) observed signi cantly worse (metabolic) health outcomes in East German than West German young women, but not for men. Similar to the results presented in this study, early-life experiences of distress in the family environment might drive these adverse health outcomes for women. Scarred for Life • 719 Our  ndings on the relevance of parental unemployment during an experience of an economic crisis suggest an accumulation of mental health risks from infancy to young adulthood rooted in economic stress (cf. Lindström et al. 2012, 2014). As previous research has unambiguously shown, the health hazard for children in settings of economic stress, both in the short- and in the long-term (Lindström et al. 2012; Moustgaard et al. 2018; Nikolova/Nikolaev 2021), is nested in their families (cf. family stress model, Masarik/Conger 2017). Our investigation of the family experience of the economic crisis, i.e., parental unemployment, showed negative associations with early-adult mental health, corresponding to insights from recent research on health consequences of childhood economic crises (cf. Briody 2021; Cui/Zack 2013; Rajmil et al. 2013) and on children’s immediate mental health and long-term stress dysregulation consequences of experiencing economic hardship (Evans/Kim 2007; Solantaus et al. 2004). Unlike literature assessing longer-term physical health effects of early-life exposure to economic stress (cf. Alessie et al. 2019; van den Berg et al. 2011), our results show little to no evidence for physical health effects of the post-reuni cation economic crisis in East Germany. We should acknowledge the nature of our chosen health outcome: the SF-12 assesses physical and mental health functioning but not its causes (RAND Corporation 2020). Subsequently, we might underestimate the actual effects of the economic crisis on physical and mental health, as only their expression in health functioning was analysed. Concerning the young age of our sample, we might expect more far-reaching health consequences in the longer term. However, we also found signi cantly lower physical and mental health among East German men and women who migrated to West Germany during the early 1990s compared to West German respondents who did not migrate. This result might indicate that the initial economic shock of the reuni cation in East Germany explains the adverse early-adult health outcomes of those affected better than the consequent economic crisis itself. Hence, the experience of the economic and political shock in East Germany accompanying the German reuni cation might be the common denominator in both East German groups, those who did migrate to West Germany and those who remained in East Germany. On the other hand, we acknowledge that the adaptation process that families had to go through who migrated from East to West Germany might have negatively contributed to the physical and mental health of their children, and potentially have cancelled out the health bene ts gained by withdrawing from the economic crisis though moving to West Germany. Prior studies provide evidence for adverse physical and mental health effects of residential relocation experienced during early life (cf. Simsek et al. 2021; Tseliou et al. 2016) and we cannot exclude the existence of additional stressors related to the migration to West Germany (cf. Kley 2013; Rosenbaum-Feldbrügge et al. 2022). At the same time, the relocation to West Germany implied the escape from the economic crisis in East Germany that potentially led to a decline in economic stress in the parents and had an alleviating effect on the children. In addition, a positive health selection was observed among East-West German migrants (cf. Westphal 2016) indicating that the respective • Lara Bister, Jeroen Spijker, Fanny Janssen, Tobias Vogt720 children and adolescents likely belonged to families that were able to mitigate negative consequences of the move. 5.3 Strengths and limitations Our study is the  rst to investigate the association between economic crises during infancy, childhood and adolescence and early-adult physical and mental health in post-reuni cation East Germany. We paid particular attention to the role of family stress by assessing parental unemployment during the economic crisis. Our data source, the data link between the German Socio-Economic Panel (SOEP) survey and social security pension records (SOEP-RV), allows us to identify the exposure to the economic crisis studied (i.e., the family household residence when the initial economic shock happened right after the German reuni cation in 1990 as well as throughout the economic crisis from 1990 to 1994), to reconstruct economic family conditions during the economic crisis through linked parental unemployment information, and to obtain health information of our respondents in early adulthood. We are among the  rst to utilise this new and unique linkage between the SOEP and the pension registers. Furthermore, given that West Germany did not experience the post-reuni cation economic crisis, we can utilise the quasi-experimental situation of the East German respondents experiencing a treatment with the West German respondents serving as a control group (under the application of advanced methods of sample weighting). Nonetheless, our study has some data-driven limitations. First, we used the SF-12 score for physical and mental health functioning, a self-reported health measure. The SF-12 is calculated following a validated and population-based algorithm (Andersen et al. 2007), which may lead to lacking comparability with other studies. We further acknowledge potential biases in our results due to cultural differences in health perceptions between East and West Germany (Lampert et al. 2019) and – probably to a lesser extent due to the young age of our study population – inherently different public health systems in pre- reuni cation East and West Germany (Busse/Riesberg 2005; Lampert et al. 2019). On the other hand, the converging cultural and living conditions as well as having spent a large part of their lives in reuni ed Germany may also have led to further convergence in the answers to the health questions in the SOEP between the East and the West German respondents. Accordingly, East Germans might – to a certain extent – have underreported their health-related restrictions, potentially leading to an underestimation of the health consequences of the post-reuni cation economic crisis for East Germans (cf. Nolte/McKee 2004). Second, the SOEP does not consider all of the potential effects of the post- reuni cation economic crisis in East Germany, including developmental information during the post-reuni cation economic crisis on all cohorts we are observing. Since the SOEP only includes individual respondents from age 17 onwards, we faced a substantial monitoring gap between exposure to the economic crisis and measuring the health outcomes. Subsequently, our results may suffer unobserved heterogeneity (cf. Schmiedeberg/Bozoyan 2021), and the observed effect of the Scarred for Life • 721 economic crisis may be partially underestimated due to prevailing East-West German health perception differences. Furthermore, we potentially underestimated the health impact over the years due to the reuni cation gains in East Germany, i.e., improved economic perspective for younger generations (cf. Forkel/Silbereisen 2001). While we cannot analytically consider these aspects, they should be kept in mind when interpreting our results. Although we considered the exposure to parental unemployment, the experience of this economic crisis may have varied enormously with the family’s socio- economic status, e.g., parental education. As previous studies have shown, children in families with higher education show fewer adverse health effects of an economic crisis (Rajmil et al. 2013). Unfortunately, we were unable to consider parental education due to many missing values in the respective measures (see Table A1). We approximated the socio-economic background by adding the mothers’ age at the child’s birth, yet, acknowledge that the omission of parental education might have led to an underestimation of the adverse health effects of exposure to the economic crisis. 5.4 Conclusion and implications Despite some limitations, we conclude that, in line with our hypotheses, the exposure to the post-reuni cation economic crisis in East Germany during infancy, childhood and adolescence was associated with lower mental health for women, especially for those who experienced parental unemployment during the economic crisis period. We thereby emphasise the relevance of this direct exposure to severe economic stress for girls regarding their later-life mental health. Although we did not observe any statistically signi cant effects for men (although most point estimates were below zero), we cannot exclude an in uence of severe early-life economic stress in boys (e.g., on other health outcomes such as health behaviour or behavioural problems). Nevertheless, our  ndings deliver insights into the long-term health development of the currently understudied East German reuni cation cohorts. In addition, our  ndings suggest long-term health disadvantages for those experiencing severe economic stress during early life, implying a double burden to these populations. Future research should therefore follow the East German reuni cation cohorts to monitor their health development throughout their adulthood, as the early- adult health disadvantages may translate into worse physical and mental health at older ages. On a more general note, our results once again urge policy makers to especially focus on the support of directly affected families in times of economic crisis and implicate early prevention methods for children and adolescents exposed to these economic adversities protecting their long-term mental health. Acknowledgements This study originates from the  rst author’s (Lara Bister) master thesis in the prorgamme of the European Doctoral School of Demography 2019/20. The authors • Lara Bister, Jeroen Spijker, Fanny Janssen, Tobias Vogt722 thank all members of the Centre d’Estudis Demogrà cs, Universitat Autònoma de Barcelona and especially Diederik Bortien whose ideas helped to develop the study. Lara Bister further acknowledges the International Max Planck Research School for Population, Health and Data Science for the training and support in conducting this research. Fanny Janssen has received funding from the Netherlands Organisation for Scienti c Research (NWO) (grant no. VIC.191.019). References Adler, Marina A. 1997: Social Change and Declines in Marriage and Fertility in Eastern Germany. In: Journal of Marriage and the Family 59,1: 37-49. https://doi.org/10.2307/353660 Alastalo, Hanna et al. 2013: Early life stress and blood pressure levels in late adulthood. In: Journal of Human Hypertension 27,2: 90-94. https://doi.org/10.1038/jhh.2012.6 Alessie, Rob J. M. et al. 2019: Economic conditions at birth and cardiovascular disease risk in adulthood: Evidence from post-1950 cohorts. In: Social Science & Medicine 224: 77-84. https://doi.org/10.1016/j.socscimed.2019.02.006 Anagnostopoulos, Dimitris C.; Soumaki, Eugenia 2013: The state of child and adolescent psychiatry in Greece during the international  nancial crisis: a brief report. In: European Child & Adolescent Psychiatry 22,2: 131-134. https://doi.org/10.1007/s00787-013-0377-y Andersen, Hanfried H. et al. 2007: Computation of Standard Values for Physical and Mental Health Scale Scores Using the SOEP Version of SF12v2. In: Ferrer-i-Carbonell, Ada; Grabka, Markus M.; Kroh, Martin (Eds.): Schmollers Jahrbuch: Journal of Applied Social Science Studies 127,1: 171-182. Ben-Shlomo, Yoav; Mishra, Gita; Kuh, Diana 2014: Life Course Epidemiology. In: Ahrens, Wolfgang; Pigeot, Iris (Eds.): Handbook of Epidemiology. New York: Springer: 1521- 1549. https://doi.org/10.1007/978-0-387-09834-0_56 Berth, Hendrik et al. 2011: Arbeitslosigkeit und Gesundheit – Ergebnisse der Sächsischen Längsschnittstudie. In: Erwerbslosigkeit: Handlungsansätze zur Gesundheitsförderung. Lengerich: Papst Science Puplishers: 35-53. Bister, Lara; Janssen, Fanny; Vogt, Tobias 2022: Early-life exposure to economic stress and metabolic risks in young adulthood: the children of the reuni cation in East Germany. In: Journal of Epidemiology and Community Health 76,9: 786-791. https://doi.org/10.1136/jech-2021-218637 Brandt, Martina; Deindl, Christian; Hank, Karsten 2012: Tracing the origins of successful aging: The role of childhood conditions and social inequality in explaining later life health. In: Social Science & Medicine 74,9: 1418-1425. https://doi.org/10.1016/j.socscimed.2012.01.004 Briody, Jonathan 2021: Parental unemployment during the Great Recession and childhood adiposity. In: Social Science & Medicine 275, 113798. https://doi.org/10.1016/j.socscimed.2021.113798 Burda, Michael C.; Hunt, Jennifer A. 2001: From Reuni cation to Economic Integration: Productivity and the Labor Market in Eastern Germany. In: Brookings Papers on Economic Activity 2: 1-92. https://doi.org/10.1353/eca.2001.0016 Scarred for Life • 723 Burgess, Simon; Propper, Carol; Rigg, John 2004: The impact of low income on child health: Evidence from a birth cohort study. Research Paper No. CASE085. London: London School of Economics [http://eprints.lse.ac.uk/6305/1/The_Impact_of_Low_ Income_on_Child_Health_Evidence_from_a_Birth_Cohort_Study.pdf, 30.11.2022]. Busse, Reinhard; Riesberg, Annette 2005: Gesundheitssysteme im Wandel: Deutschland. Copenhagen: World Health Organisation [https://www.euro.who.int/__data/assets/ pdf_ le/0019/108460/E85472G.pdf, 30.11.2022]. Carslake, David et al. 2017: Associations of parental age with health and social factors in adult offspring. Methodological pitfalls and possibilities. In: Scienti c Reports 7: 45278. https://doi.org/10.1038/srep45278 Chevalier, Arnaud; Marie, Olivier 2017: Economic Uncertainty, Parental Selection, and Children’s Educational Outcomes. In: Journal of Political Economy 125,2: 393-430. https://doi.org/10.1086/690830 Chzhen, Yekaterina 2016: Perceptions of the Economic Crisis in Europe: Do Adults in Households with Children Feel a Greater Impact? In: Social Indicators Research 127: 341-360. https://doi.org/10.1007/s11205-015-0956-z Conger, Katherine Jewsbury et al. 2012: Economic Hardship and Its Consequences Across Generations. In: Maholmes, Valerie; King, Rosalind B. (Eds.): The Oxford handbook of poverty and child development. Oxford University Press: 37-53. https://doi.org/10.1093/oxfordhb/9780199769100.013.0002 Conger, Rand D.; Conger, Katherine J. 2002: Resilience in Midwestern Families: Selected Findings from the First Decade of a Prospective, Longitudinal Study. In: Journal of Marriage and Family 64,2: 361-373. https://doi.org/10.1111/j.1741-3737.2002.00361.x Costa, Diogo et al. 2020: Children mental health after the 2008 global economic crisis: Assessing the impact of austerity in Portugal. In: Children and Youth Services Review 118: 105332. https://doi.org/10.1016/j.childyouth.2020.105332 Cui, Wanjun; Zack, Matthew M. 2013: Trends in Health-Related Quality of Life Among Adolescents in the United States, 2001-2010. In: Preventing Chronic Disease 10: 120334. https://doi.org/10.5888/pcd10.120334 Currie, Janet; Stabile, Mark 2002: Socioeconomic Status and Health: Why is the Relationship Stronger for Older Children? NBER Working Paper No. 9098. Cambridge: National Bureau of Economic Research [https://www.nber.org/papers/w9098.pdf, 30.11.2022]. Dedovic, Katarina et al. 2009: The role of sex and gender socialization in stress reactivity. In: Developmental Psychology 45,1: 45-55. https://doi.org/10.1037/a0014433 Doyle, Orla; Harmon, Colm P.; Walker, I. 2007: Impact of Parental Income and Education on Child Health: Further Evidence for England. Discussion Paper No. WP/6/2007. Dublin: Geary Instititute, University College Dublin. Easterlin, Richard A.; Plagnol, Anke C. 2008: Life satisfaction and economic conditions in East and West Germany pre- and post-uni cation. In: Journal of Economic Behavior & Organization 68,3-4: 433-444. https://doi.org/10.1016/j.jebo.2008.06.009 Elder, Glen H. Jr. 1998: The Life Course as Developmental Theory. In: Child Development 69,1: 1-12. https://doi.org/10.2307/1132065 Elder, Glen H. Jr. 1999: Children of the great depression: social change in life experience (25th anniversary ed.). Boulder, Colo: Westview Press. Elder, Glen H. Jr.; Caspi, Avshalom 1988: Economic Stress in Lives: Developmental Perspectives. In: Journal of Social Issues 44,4: 25-45. https://doi.org/10.1111/j.1540-4560.1988.tb02090.x • Lara Bister, Jeroen Spijker, Fanny Janssen, Tobias Vogt724 Evans, Gary W.; Kim, Pilyoung 2007: Childhood Poverty and Health: Cumulative Risk Exposure and Stress Dysregulation. In: Psychological Science 18,11: 953-957. https://doi.org/10.1111/j.1467-9280.2007.02008.x Federal Statistical Of ce 2020a: Bevölkerungszahlen [https://www-genesis.destatis.de/ genesis/online?operation=themes&levelindex=0&levelid=1611398061266&code=12# abreadcrumb, 10.05.2023]. Federal Statistical Of ce 2020b: Arbeitsmarktstatistik der Bundesagentur für Arbeit [https://www-genesis.destatis.de/genesis/online?operation=abruftabelleBearbeiten &levelindex=1&levelid=1603461532058&auswahloperation=abruftabelleAuspraegu ngAuswaehlen&auswahlverzeichnis=ordnungsstruktur&auswahlziel=werteabruf&co de=13211-0001&auswahltext=&werteabruf=Werteabruf#abreadcrumb, 10.05.2023]. Forkel, Ines; Silbereisen, Rainer K. 2001: Family Economic Hardship and Depressed Mood Among Young Adolescents From Former East and West Germany. In: American Behavioral Scientist 44,11: 1955-1971. https://doi.org/10.1177/00027640121958131 Frijters, Paul; Haisken-DeNew, John P.; Shields, Michael A. 2005: The causal effect of income on health: Evidence from German reuni cation. In: Journal of Health Economics 24,5: 997-1017. https://doi.org/10.1016/j.jhealeco.2005.01.004 Gard, Arianna M. et al. 2020: Evaluation of a longitudinal family stress model in a population-based cohort. In: Social Development 29,4: 1155-1175. https://doi.org/10.1111/sode.12446 Ge, Yinjian et al. 2019: Association of physical activity, sedentary time, and sleep duration on the health-related quality of life of college students in Northeast China. In: Health and Quality of Life Outcomes 17,1: 124. https://doi.org/10.1186/s12955-019-1194-x Glei, Dana A. et al. 2007: Do Chronic Stressors Lead to Physiological Dysregulation? Testing the Theory of Allostatic Load. In: Psychosomatic Medicine 69,8: 769-776. https://doi.org/10.1097/PSY.0b013e318157cba6 Grigoriev, Pavel; Pechholdová, Markéta 2017: Health Convergence Between East and West Germany as Re ected in Long-Term Cause-Speci c Mortality Trends: To What Extent was it Due to Reuni cation? In: European Journal of Population 33,5: 701-731. https://doi.org/10.1007/s10680-017-9455-z Hainmueller, Jens 2012: Entropy Balancing for Causal Effects: A Multivariate Reweighting Method to Produce Balanced Samples in Observational Studies. In: Political Analysis 20,1: 25-46. https://doi.org/10.1093/pan/mpr025 Hainmueller, Jens; Xu, Yiqing 2013: Ebalance: A Stata Package for Entropy Balancing. In: Journal of Statistical Software 54,7. https://doi.org/10.2139/ssrn.1943090 Hauser, Richard 1996: Die Entwicklung der Einkommensverteilung in den neuen Bundesländern seit der Wende. In: Diewald, Martin; Mayer, Karl Ulrich (Eds.): Zwischenbilanz der Wiedervereinigung. Wiesbaden: VS Verlag für Sozialwissenschaften: 165-188. https://doi.org/10.1007/978-3-322-95728-3_8 Heiland, Frank W. 2004: Trends in East-West German Migration from 1989 to 2002. In: Demographic Research 11,7: 173-194. https://doi.org/10.4054/DemRes.2004.11.7 Hessel, Philipp; Vandoros, Sotiris; Avendano, Mauricio Pabon 2014: The differential impact of the  nancial crisis on health in Ireland and Greece: a quasi-experimental approach. In: Public Health 128,10: 911-919. https://doi.org/10.1016/j.puhe.2014.08.004 Hofer, Manfred et al. 1995: Der soziale Wandel aus Sicht ost- und westdeutscher Familien, psychisches Wohlbe nden und autoritäre Vorstellungen. In: Nauck, Bernhard; Norbert F. Schneider; Tölke, Angelika (Eds.): Familie und Lebensverlauf im gesellschaftlichen Umbruch. Stuttgart: F. Enke. Scarred for Life • 725 Huinink, Johannes; Kreyenfeld, Michaela 2006: Family formation in times of abrupt social and economic change. In: Huinink, Johannes; Kreyenfeld, Michaela (Eds.): After the fall of the wall: Life courses in the transformation of East Germany. Stanford: Stanford University Press: 170-190. Jörngården, Anna; Wettergen, Lena; von Essen, Louise 2006: Measuring health-related quality of life in adolescents and young adults: Swedish normative data for the SF-36 and the HADS, and the in uence of age, gender, and method of administration. In: Health and Quality of Life Outcomes 4,91. https://doi.org/10.1186/1477-7525-4-91 Karanikolos, Marina et al. 2016: Effects of the Global Financial Crisis on Health in High- Income Oecd Countries: A Narrative Review. In: International Journal of Health Services 46,2: 208-240. https://doi.org/10.1177/0020731416637160 Kley, Stefanie 2013: Migration in the Face of Unemployment and Unemployment Risk: a Case Study of Temporal and Regional Effects. In: Comparative Population Studies 38,1: 109-136. https://doi.org/10.12765/CPoS-2013-04 Knies, Gundi; Spiess, C. Katharina 2007: Regional Data in the German Socio- Economic Panel Study (SOEP). Data Documentation 17. Berlin: DIW Deutsches Institut für Wirtschaftsforschung [https://www.diw.de/documents/publikationen/73/ diw_01.c.55738.de/diw_datadoc_2007-017.pdf, 10.05.2023]. Kuh, Diana et al. 2013: Life course epidemiology, ageing research, and maturing cohort studies: a dynamic combination for understanding healthy ageing. In: Kuh, Diana et al. (Eds.): A Life Course Approach to Healthy Ageing. Oxford University Press: 3-15. https://doi.org/10.1093/acprof:oso/9780199656516.003.0001 Lampert, Thomas et al. 2019: Beschreibung der gesundheitlichen Lage in Ost und West sowie deren Entwicklung in den letzten 30 Jahren. In: Journal of Health Monitoring 4,S2: 1-25. https://doi.org/10.25646/6076 Liebig, Stefan et al. 2019: Socio-Economic Panel (SOEP), data from 1984-2018. In: SOEP Socio-Economic Panel Study. https://doi.org/10.5684/SOEP-CORE.V35 Lindström, Martin; Fridh, Maria; Rosvall, Maria 2014: Economic stress in childhood and adulthood, and poor psychological health: Three life course hypotheses. In: Psychiatry Research 215,2: 386-393. https://doi.org/10.1016/j.psychres.2013.11.018 Lindström, Martin; Hansen, Kristina; Rosvall, Maria 2012: Economic stress in childhood and adulthood, and self-rated health: a population based study concerning risk accumulation, critical period and social mobility. In: BMC Public Health 12,761. https://doi.org/10.1186/1471-2458-12-761 Lüthen, Holger et al. 2022: SOEP-RV: Linking German Socio-Economic Panel Data to Pension Records. In: Jahrbücher für Nationalökonomie und Statistik 242,2: 291-307. https://doi.org/10.1515/jbnst-2021-0020 Masarik, April S.; Conger, Rand D. 2017: Stress and child development: a review of the Family Stress Model. In: Current Opinion in Psychology 13: 85-90. https://doi.org/10.1016/j.copsyc.2016.05.008 Moscarelli, Massimo; Manning, Willard 2009: Adults respondent with a problem drinking father were more likely to have been diagnosed with mental health problems relative to other respondents, while those with a problem drinking mother had poor self-perceived health and mental health (SF-12) scores. In: The journal of mental health policy and economics 12,2: 53-54. Moustgaard, Heta; Avendano, Mauricio; Martikainen, Pekka 2018: Parental Unemployment and Offspring Psychotropic Medication Purchases: A Longitudinal Fixed-Effects Analysis of 138,644 Adolescents. In: American Journal of Epidemiology 187,9: 1880-1888. https://doi.org/10.1093/aje/kwy084 • Lara Bister, Jeroen Spijker, Fanny Janssen, Tobias Vogt726 Mucci, Nicola et al. 2016: The correlation between stress and economic crisis: a systematic review. In: Neuropsychiatric Disease and Treatment 12: 983-993. https://doi.org/10.2147/NDT.S98525 Nikolova, Milena; Nikolaev, Boris N. 2018: Family matters: The effects of parental unemployment in early childhood and adolescence on subjective well-being later in life. In: Journal of Economic Behavior & Organization 181: 312-331. https://doi.org/10.1016/j.jebo.2018.05.005 Nolte, Ellen; McKee, Martin 2004: Changing health inequalities in east and west Germany since uni cation. In: Social Science & Medicine 58,1: 119-136. https://doi.org/10.1016/S0277-9536(03)00153-9 Powdthavee, Nattavudh; Vernoit, James 2013: Parental unemployment and children’s happiness: A longitudinal study of young people’s well-being in unemployed households. In: Labour Economics 24: 253-263. https://doi.org/10.1016/j.labeco.2013.09.008 Rajmil, Luis et al. 2014: Impact of the 2008 Economic and Financial Crisis on Child Health: A Systematic Review. In: International Journal of Environmental Research and Public Health 11,6: 6528-6546. https://doi.org/10.3390/ijerph110606528 Rajmil, Luis et al. 2013: Impact of the economic crisis on children’s health in Catalonia: a before-after approach. In: BMJ Open 3,8: e003286. https://doi.org/10.1136/bmjopen-2013-003286 RAND Corporation 2020: 12-Item Short Form Survey (SF-12) [https://www.rand.org/ health-care/surveys_tools/mos/12-item-short-form.html, 27.10.2020]. Reibling, Nadine et al. 2017: Depressed during the depression: has the economic crisis affected mental health inequalities in Europe? Findings from the European Social Survey (2014) special module on the determinants of health. In: European Journal of Public Health 27: 47-54. https://doi.org/10.1093/eurpub/ckw225 Rios-Avila, Fernando; Maroto, Michelle Lee 2022: Moving Beyond Linear Regression: Implementing and Interpreting Quantile Regression Models With Fixed Effects. In: Sociological Methods & Research. https://doi.org/10.1177/00491241211036165 Rodrigues, Daniela et al. 2021: The Great Recession weighted on Portuguese children: A structural equation modeling approach considering eating patterns. In: American Journal of Human Biology 34,5: e23692. https://doi.org/10.1002/ajhb.23692 Rosenbaum-Feldbrügge, Matthias; Stawarz, Nico; Sander, Nikola 2022: 30 Years of East- West Migration in Germany: A Synthesis of the Literature and Potential Directions for Future Research. In: Comparative Population Studies 47. https://doi.org/10.12765/CPoS-2022-08 Schmiedeberg, Claudia; Bozoyan, Christiane 2021: Do Economic Hardship and Pressure Really In uence Parenting? In: European Sociological Review 37,2: 287-304. https://doi.org/10.1093/esr/jcaa051 Simsek, Müge; Costa, Rafael; de Valk, Helga A. G. 2021: Childhood residential mobility and health outcomes: A meta-analysis. In: Health & Place 71: 102650. https://doi.org/10.1016/j.healthplace.2021.102650 Solantaus, Tytti; Leinonen, Jenni; Punamäki, Rija-Leena 2004: Children’s Mental Health in Times of Economic Recession: Replication and Extension of the Family Economic Stress Model in Finland. In: Developmental Psychology 40,3: 412-429. https://doi.org/10.1037/0012-1649.40.3.412 The Conference Board Total Economy Database™ 2013: Labor and Labor Productivity Country Details, 1950-2013 [http://www.conference-board.org/data/ economydatabase/, 10.05.2023]. Scarred for Life • 727 Thompson, Kristina; Ophem, Johan van; Wagemakers, Annemarie 2019: Studying the impact of the Eurozone’s Great Recession on health: Methodological choices and challenges. In: Economics & Human Biology 35: 162-184. https://doi.org/10.1016/j.ehb.2019.06.004 Tseliou, Foteini et al. 2016: The impact of childhood residential mobility on mental health outcomes in adolescence and early adulthood: a record linkage study. In: Journal of Epidemiology and Community Health 70,3: 278-285. https://doi.org/10.1136/jech-2015-206123 Turner-Bowker, Diane; Hogue, Sarah J. 2014: Short Form 12 Health Survey (SF-12). In: Michalos, Alex C. (Ed.): Encyclopedia of Quality of Life and Well-Being Research. Dordrecht: Springer Netherlands: 5954-5957. https://doi.org/10.1007/978-94-007-0753-5_2698 van den Berg, Gerard J.; Doblhammer-Reiter, Gabriele; Christensen, Kaare 2011: Being Born Under Adverse Economic Conditions Leads to a Higher Cardiovascular Mortality Rate Later in Life: Evidence Based on Individuals Born at Different Stages of the Business Cycle. In: Demography 48,2: 507-530. https://doi.org/10.1007/s13524-011-0021-8 van den Berg, Gerard J.; Lindeboom, Maarten; Portrait, France 2006: Economic Conditions Early in Life and Individual Mortality. In: American Economic Review 96,1: 290-302. https://doi.org/10.1257/000282806776157740 van Roode, Thea et al. 2017: Life-Course Relationship between Socioeconomic Circumstances and Timing of First Birth in a Birth Cohort. In: PLOS ONE 12,1: e0170170. https://doi.org/10.1371/journal.pone.0170170 Vogt, Tobias C. 2013: How Many Years of Life Did the Fall of the Berlin Wall Add? A Projection of East German Life Expectancy. In: Gerontology 59,3: 276-282. https://doi.org/10.1159/000346355 Wagner, Gert G.; Frick, Joachim R.; Schupp, Jürgen 2007: The German Socio-Economic Panel Study (SOEP): Scope, Evolution and Enhancements. In: Schmollers Jahrbuch: Journal of Applied Social Science Studies 127,1: 139-169. Westphal, Christina 2016: Healthy Migrants? Health Selection of Internal Migrants in Germany. In: European Journal of Population 32,5: 703-730. https://doi.org/10.1007/s10680-016-9397-x Widding-Havneraas, Tarjei; Pedersen, Siri Hansen 2020: The role of welfare regimes in the relationship between childhood economic stress and adult health: a multilevel study of 20 European countries. In: SSM – Population Health 12: 100674. https://doi.org/10.1016/j.ssmph.2020.100674 • Lara Bister, Jeroen Spijker, Fanny Janssen, Tobias Vogt728 Date of submission: 08.03.2022 Date of acceptance: 29.08.2023 Lara Bister (). University of Groningen, Faculty of Spatial Sciences. Groningen, The Netherlands. E-mail: l.bister@rug.nl URL: https://www.rug.nl/staff/l.bister/?lang=en Dr. Jeroen Spijker. Centre d’Estudis Demogrà cs, Universitat Autònoma de Barcelona, Bellaterra, Spain. E-mail: jspijker@ced.uab.es URL: https://ced.cat/en/directori/jeroen-spijker/ Prof. Dr. Fanny Janssen. University of Groningen, Netherlands Interdisciplinary Demographic Institute – KNAW. Den Haag, The Netherlands. E-mail: janssen@nidi.nl URL: https://nidi.nl/en/employees/fanny-janssen/ Dr. Tobias Vogt. University of Groningen, Population Research Centre, Faculty of Spatial Sciences. Groningen, The Netherlands. Prasanna School of Public Health, Manipal Academy for Higher Education. Karnataka, India. E-mail: t.c.vogt@rug.nl URL: https://research.rug.nl/en/persons/tobias-vogt Scarred for Life • 729 Appendix Fig. A1: Sample selection of the used sample in the analysis derived from the SOEP population (active respondents only) Source: German Socio-Economic Panel (SOEP) and SOEP-RV link • Lara Bister, Jeroen Spijker, Fanny Janssen, Tobias Vogt730 Fi g. A 2: R es ul ts s en si tiv it y an al ys is u si ng (r es po n d en t- re po rt ed ) p ar en ta l h ig he st e d uc at io n as a p ro xy fo r th e so ci o ec o n o m ic s ta tu s in st ea d o f m o th er s’ a ge N o te : R ep lic at io n o f m ai n an al ys is ( le ft c o lu m n ) u si ng ( re sp o n d en t- re p o rt ed ) m ea su re f o r hi g he st p ar en ta l e d uc at io n (r ig h t co lu m n, f ro m S O E P d at a se t b io p ar en ) (N = 2, 19 8) . M ea su re c o ns is ts o f in fo rm at io n o n th e m at er na l a nd p at er na l s ch o o l a n d p ro fe ss io na l e d u ca tio n d eg re es , c o m b in ed in li ne w ith IS C E D -b as ed c at eg o ri es : ( 1) n o fo rm al e d uc at io n, (2 ) l o w er s ec o nd ar y ed uc at io n, (3 ) u p p er s ec o n d ar y ed uc at io n, (4 ) p o st -s ec o nd ar y ed uc at io n, o r (5 ) t er tia ry e d uc at io n . D is tr ib ut io n o f h ig h es t p ar en ta l e d uc at io n : n= 84 r es p o n d en ts (3 .8 p er ce nt ) h ad p ar en ts w ith a s ec o nd ar y sc ho o l e d uc at io n o n ly , n = 1, 78 6 (8 1. 3 p er ce nt ) w ith p o st -s ec o n d ar y (n o n -a ca d em ic ) p ro fe ss io na l ed u ca tio n a n d n = 32 8 (1 4. 9 p er ce nt ) w ith t er tia ry ( ac ad em ic ) p ro fe ss io n al ed uc at io n. S o u rc e: G er m an S o ci o -E co n o m ic P an el (S O E P ) a n d S O E P- R V li nk Scarred for Life • 731 Tab. A1: Missing values in highest parental education variable (SOEP data set pl) Highest parental West Germany East Germany Total education (ISCED) ISCED 1 9 0.66% 0 0.00% 9 0.39% ISCED 2 4 0.30% 0 0.00% 4 0.17% ISCED 3 1 0.07% 1 0.10% 2 0.09% ISCED 4 154 11.37% 259 26.37% 413 17.67% ISCED 5 30 2.21% 18 1.83% 48 2.05% ISCED 6 0 0.00% 3 0.31% 3 0.13% Missing 1,157 85.39% 701 71.38% 1,858 79.50% Total 1,355 982 2,337 Note: percentages show column percentages, ISCED levels are in accordance with ISCEDv1997. Source: German Socio-Economic Panel (SOEP) and SOEP-RV link • Lara Bister, Jeroen Spijker, Fanny Janssen, Tobias Vogt732 Ta b . A 2: M is si n g va lu es in h ig h es t p ar en ta l e d uc at io n va ri ab le (S O EP d at a se t p l) S am pl e g ro u p: Tr ea tm en t C o n tr o l ( b ef o re m at ch in g) C o n tr o l ( af te r m at ch in g) M at ch in g va ri ab le s: M ea n V ar ia n ce S ke w n es s M ea n V ar ia n ce S ke w n es s M ea n V ar ia n ce S ke w n es s E as t- W es t m ig ra ti o n 0. 29 7 0. 20 9 0. 8 87 0. 26 1 0. 19 3 1. 08 7 0. 29 7 0. 20 9 0. 8 87 P ar en ta l u n em pl o ym en t 0. 13 0 0. 03 6 1. 67 2 0. 05 4 0. 02 1 4. 0 51 0. 13 0 0. 06 8 2. 2 94 B ir th y ea r 19 8 2 19 .9 3 -0 .3 01 19 8 2 20 .6 3 -0 .3 67 19 8 2 19 .8 -0 .3 92 Y ea r o f S F- 12 20 03 4. 23 2. 22 6 20 03 5. 21 3 2. 4 4 5 20 03 4. 5 39 2. 5 37 G en d er : fe m al e 0. 4 8 9 0. 25 0 0. 04 5 0. 50 9 0. 25 0 -0 .0 34 0. 4 8 9 0. 25 0 0. 04 5 N u m b er o f si b lin g s 1. 26 9 1. 13 7 1. 6 0 6 1. 6 47 1. 83 4 1. 60 7 1. 27 1. 0 8 6 1. 22 3 M o th er ’s a g e in y ea rs 25 .1 1 21 .7 9 0. 9 04 26 .6 27 .3 2 0. 52 25 .1 1 23 .8 0. 60 8 N ot e: E nt ro p y b al an ci ng s am p le w ei gh ts w er e ca lc ul at ed in S ta ta S E 1 7 us in g “e b al an ce ” (H ai nm ue lle r/ X u 2 01 1: E b al an ce : A S ta ta Pa ck ag e fo r E nt ro py B al an ci ng . S S R N E le ct ro ni c Jo ur n al . ht tp s: //d oi .o rg /1 0. 21 39 /s sr n. 19 43 09 0) . M at ch in g of t he t re at m en t sa m pl e (E as t G er m an b o rn i nd iv id ua ls ) an d t h e co nt ro l sa m pl e (W es t G er m an b o rn i nd iv id ua ls ) us in g m ea su re s o f Ea st -W es t m ig ra tio n, p ar en ta l u ne m pl oy m en t b et w ee n 19 90 a nd 1 99 4, b irt h ye ar , y ea r of S F- 12 a ss es sm en t, ge nd er (m al e an d fe m al e) , n um be r of s ib lin g s, m ot he r’ s ag e at b ir th . C o nv er g en ce a ch ie ve d af te r el ev en i te ra tio ns o f ba la nc in g op tim is at io n w ith m ax im um d iff er en ce b et w ee n tr ea tm en t a nd c o nt ro l g ro up s o f 0 .0 02 99 19 99 . A to ta l o f 9 82 s am p lin g w ei gh ts a ch ie ve d. S ou rc e: G er m an S o ci o -E co no m ic P an el (S O E P) a nd S O EP -R V li nk Scarred for Life • 733 Ta b . A 3. 1- A 3. 4: Q u an til e re g re ss io n ta b le s an d av er ag e m ar g in al e ff ec t t ab le s (N = 2 ,3 37 ) Ta b . A 3. 1: Q ua n til e re g re ss io n re su lt s o n S F- 12 P hy si ca l C o m p o ne nt S co re (P C S ) ( N = 2 ,3 37 ) M en W o m en Q u an ti le s Q u an ti le s V ar ia b le s 20 % 4 0% 6 0% 8 0% 20 % 4 0% 6 0% 8 0% B o rn in E as t G er m an y -0 .5 53 0. 11 0 0. 41 6 -0 .0 58 3 0. 95 4 0. 22 7 0. 61 6 0. 61 7 (r e f.: W es t G er m an y) (0 .8 03 ) (0 .7 3 0) (0 .3 60 ) (0 .3 4 6) (0 .6 3 3) (1 .0 9 9) (0 .6 95 ) (0 .5 45 ) E as t- W es t M ig ra ti o n -0 .2 21 0. 25 2 0. 19 1 -0 .0 34 4 0. 8 59 -0 .4 5 6 1. 16 8 1. 36 7* * (r e f.: n o m ig ra ti o n) (1 .0 29 ) (0 .6 18 ) (0 .4 60 ) (0 .2 81 ) (0 .7 0 9) (1 .0 37 ) (0 .7 6 8) (0 .5 87 ) B o rn in E as t G er m an y X M ig ra ti on E as t X M ig ra ti o n -0 .4 32 -0 .6 58 -1 .0 3 0* -0 .9 8 3* ** -1 .1 92 0. 73 7 -0 .8 86 -1 .7 36 * (1 .5 16 ) (0 .9 0 5) (0 .6 07 ) (0 .3 81 ) (0 .9 30 ) (1 .6 45 ) (1 .1 07 ) (0 .8 9 4) B o rn in E as t G er m an y X P ar en ta l u n em pl oy m en t (1 99 0 -9 4) W es t X U n em p ly o m en t 0. 06 57 0. 26 9 -0 .1 63 -0 .5 7 7 -0 .2 4 8 -1 .3 55 -1 .0 3 6 -0 .1 91 (1 .6 32 ) (0 .9 62 ) (0 .3 07 ) (0 .6 51 ) (0 .9 35 ) (1 .0 41 ) (1 .0 70 ) (0 .7 3 0) E as t X U n em p ly o m en t -1 .2 75 ** -1 .3 2 9* * -1 .2 6 9* ** -1 .1 21 ** * -1 .1 13 -2 .0 14 -1 .3 90 -0 .7 21 (0 .6 2 8) (0 .6 59 ) (0 .3 60 ) (0 .2 51 ) (0 .7 12 ) (1 .3 12 ) (0 .8 62 ) (0 .5 45 ) A g e in 1 9 90 (r ef .: In fa n cy (a g e 0- 2) ) E ar ly c h ild h o o d ( ag e 3 -5 ) 0. 65 0 0. 26 1 0. 07 8 6 -0 .0 34 4 0. 67 2 0. 18 4 0. 3 50 0. 19 7 (0 .9 19 ) (0 .5 55 ) (0 .4 06 ) (0 .2 56 ) (0 .6 47 ) (0 .9 52 ) (0 .6 76 ) (0 .5 6 0) C h ild h o o d (a g e 6 -1 1) 0. 81 9 0. 64 8 -0 .1 9 9 -0 .1 68 0. 71 3 0. 3 81 0. 8 37 0. 47 0 (0 .9 16 ) (0 .5 3 6) (0 .3 9 6) (0 .2 8 0) (0 .6 0 6) (0 .8 86 ) (0 .5 62 ) (0 .5 3 5) A d o le sc en ce ( ag e 12 -1 7) -0 .4 9 0 -0 .6 5 4 -0 .9 23 * -1 .1 06 ** 0. 03 25 -3 .1 07 ** 0. 3 06 -0 .2 6 0 (1 .1 71 ) (0 .7 03 ) (0 .5 3 0) (0 .4 36 ) (0 .7 61 ) (1 .4 58 ) (1 .1 51 ) (0 .6 77 ) N u m b er o f si b lin g s -0 .1 8 9 -0 .0 2 94 -0 .0 20 5 -0 .0 23 3 0. 26 7 -0 .4 28 0. 17 5 0. 19 3 (0 .2 6 3) (0 .2 03 ) (0 .0 8 68 ) (0 .1 13 ) (0 .2 0 8) (0 .3 9 6) (0 .2 42 ) (0 .1 4 8) S F- 12 M C S 0. 05 38 -0 .0 4 6 4* -0 .1 0 5* ** -0 .1 74 ** * 0. 01 41 0. 16 3* ** 0. 05 97 * -0 .0 17 0 (0 .0 38 5) (0 .0 26 2) (0 .0 17 5) (0 .0 13 6) (0 .0 27 1) (0 .0 4 0 6) (0 .0 3 41 ) (0 .0 24 5) M o th er ’s a g e (z -s td .) 0. 17 2 0. 23 1 0. 21 5* * 0. 11 8 0. 4 53 ** 0. 17 3 0. 52 8* * 0. 22 9 (0 .2 70 ) (0 .2 10 ) (0 .0 9 09 ) (0 .0 9 91 ) (0 .1 99 ) (0 .3 76 ) (0 .2 6 6) (0 .1 70 ) C on st an t 50 .4 8* ** 5 8. 9 4* ** 6 4. 62 ** * 70 .5 3* ** 55 .6 6* ** 4 4. 72 ** * 52 .1 2* ** 5 8. 8 3* ** (2 .3 3 8) (1 .5 24 ) (1 .0 9 0) (0 .8 19 ) (1 .5 80 ) (2 .3 0 9) (2 .0 68 ) (1 .4 58 ) O b se rv at io n s 1, 16 8 1, 16 8 1, 16 8 1, 16 8 1, 16 9 1, 16 9 1, 16 9 1, 16 9 N ot e: R o bu st s ta nd ar d er ro rs in p ar en th es es . S ig ni  c an ce le ve ls : ** * p< 0. 00 1, * * p< 0. 01 , * p < 0. 05 . S ou rc e: G er m an S o ci o -E co no m ic P an el (S O E P) a nd S O EP -R V li nk • Lara Bister, Jeroen Spijker, Fanny Janssen, Tobias Vogt734 Ta b . A 3. 2: A ve ra ge m ar g in al e ff ec t ta b le s fo r S F- 12 P hy si ca l C o m p o ne n t S co re (P C S ) ( N = 2 ,3 37 ) M en W o m en Q u an ti le s Q u an ti le s A ve ra g e m ar g in al e ff ec ts 20 % 4 0% 6 0% 8 0% 20 % 4 0% 6 0% 8 0% B o rn in E as t G er m an y -1 .1 77 -0 .6 75 -0 .2 9 6 -0 .5 4 9* 0. 18 8 0. 21 0 -0 .1 14 0. 09 62 (r e f.: W es t G er m an y) (0 .7 4 4) (0 .4 9 5) (0 .2 37 ) (0 .3 03 ) (0 .7 4 9) (0 .6 0 6) (0 .4 19 ) (0 .2 87 ) B o rn in E as t G er m an y X M ig ra ti o n to W es t G er m an y N o m ig ra ti o n -1 .0 4 9 -0 .4 81 0. 00 65 9 -0 .2 6 0 -0 .0 34 0 0. 47 7 0. 4 08 0. 3 69 (0 .7 9 8) (0 .6 14 ) (0 .2 61 ) (0 .3 57 ) (0 .8 89 ) (0 .6 82 ) (0 .4 63 ) (0 .3 77 ) M ig ra ti o n -1 .4 82 -1 .1 3 9 -1 .0 23 -1 .2 4 3* * 0. 70 3 -0 .4 0 9 -1 .3 28 -0 .5 3 9 (1 .4 05 ) (0 .7 12 ) (0 .5 3 8) (0 .3 30 ) (1 .3 85 ) (0 .9 97 ) (0 .8 03 ) (0 .3 9 8) B o rn in E as t G er m an y X A n y p ar en ta l u n em p lo ym en t 19 9 0 -1 99 4 N o u n em p lo ym en t -0 .6 8 0 -0 .0 8 32 0. 11 3 -0 .3 47 0. 4 4 8 0. 3 50 0. 09 56 0. 57 1 (0 .6 98 ) (0 .5 81 ) (0 .3 18 ) (0 .2 79 ) (0 .8 77 ) (0 .5 5 8) (0 .4 63 ) (0 .3 51 ) U n em p lo ym en t -2 .0 21 -1 .6 82 -0 .9 93 ** -0 .8 91 -0 .2 10 -0 .0 03 4 3 -0 .4 34 -0 .6 28 (1 .6 0 4) (0 .9 6 9) (0 .3 49 ) (0 .6 4 5) (1 .4 02 ) (1 .2 71 ) (0 .7 9 0) (0 .5 19 ) O b se rv at io n s 1, 16 8 1, 16 8 1, 16 8 1, 16 8 1, 16 9 1, 16 9 1, 16 9 1, 16 9 N ot e: R o bu st s ta nd ar d er ro rs in p ar en th es es . S ig ni  c an ce le ve ls : ** * p< 0. 00 1, * * p< 0. 01 , * p < 0. 05 . S ou rc e: G er m an S o ci o -E co no m ic P an el (S O E P) a nd S O EP -R V li nk Scarred for Life • 735 Ta b . A 3. 3: Q u an til e re g re ss io n re su lt s o n S F- 12 M en ta l C o m po n en t S co re (M C S ) ( N = 2 ,3 37 ) M en W o m en Q u an ti le s Q u an ti le s V ar ia b le s 20 % 4 0% 6 0% 8 0% 20 % 4 0% 6 0% 8 0% B o rn in E as t G er m an y -2 .1 91 -1 .9 19 -2 .7 69 ** * -2 .2 98 ** * -2 .4 68 -2 .0 0 4 -1 .4 2 9 -2 .4 4 4* (r e f.: W es tG er m an y) (1 .8 97 ) (1 .2 6 6) (0 .9 25 ) (0 .8 82 ) (1 .5 11 ) (1 .4 8 6) (0 .9 51 ) (1 .0 57 ) E as t- W es t M ig ra ti o n (r ef .: n o m ig ra ti o n) -1 .7 93 -0 .7 55 -1 .3 72 -0 .7 21 -0 .6 51 -0 .0 74 0 -0 .3 22 -1 .1 08 (2 .5 74 ) (1 .1 76 ) (0 .8 9 9) (0 .4 54 ) (1 .8 56 ) (1 .3 58 ) (0 .9 6 8) (1 .0 64 ) B o rn in E as t G er m an y X M ig ra ti o n E as t X M ig ra ti o n 0. 3 69 2. 22 4 1. 61 5 2. 52 0* * -0 .2 8 0 -1 .1 90 -1 .3 90 -0 .1 13 (3 .1 3 9) (1 .8 95 ) (1 .3 8 8) (1 .0 45 ) (2 .2 74 ) (2 .2 69 ) (1 .4 0 4) (1 .4 91 ) B o rn in E as t G er m an y X P ar en ta l u n em pl oy m en t (1 99 0 -9 4) W es t X U n em p ly o m en t 0. 61 7 1. 7 74 1. 32 1 -0 .2 39 -3 .2 13 * 0. 28 0 0. 70 8 -0 .0 56 3 (2 .5 4 9) (1 .6 8 6) (1 .3 4 3) (0 .4 78 ) (1 .8 11 ) (1 .9 20 ) (1 .3 60 ) (1 .2 17 ) E as t X U n em p ly o m en t 0. 63 7 1. 13 2 0. 50 7 1. 82 0* -1 .8 12 -2 .3 33 -2 .6 37 ** -1 .1 42 (1 .7 56 ) (1 .3 61 ) (0 .9 4 8) (0 .9 61 ) (1 .3 07 ) (1 .4 93 ) (1 .1 6 6) (0 .9 11 ) A g e in 1 9 90 (r ef .: In fa n cy (a g e 0- 2) ) E ar ly c h ild h o o d ( ag e 3 -5 ) -0 .2 77 0. 3 4 4 1. 53 0* 0. 3 65 1. 63 6 -2 .0 6 0 -2 .1 21 ** -1 .2 8 8 (1 .8 92 ) (1 .0 57 ) (0 .8 4 3) (0 .6 13 ) (1 .6 14 ) (1 .4 4 9) (0 .9 5 4) (0 .9 27 ) C h ild h o o d (a g e 6 -1 1) -4 .2 3 4* * -3 .5 8 4* ** -1 .0 8 8 -1 .9 78 ** * -1 .3 28 -2 .6 16 * -1 .9 72 ** -2 .2 51 ** * (1 .7 36 ) (1 .1 33 ) (0 .8 01 ) (0 .4 70 ) (1 .5 9 6) (1 .4 75 ) (0 .7 67 ) (0 .7 20 ) A d o le sc en ce ( ag e 12 -1 7) -2 .9 46 -1 .5 10 -1 .0 0 0 -2 .1 07 ** * -1 .9 17 -4 .4 13 ** * -3 .6 03 ** * -1 .6 20 (2 .2 58 ) (1 .2 71 ) (1 .1 11 ) (0 .5 21 ) (1 .8 53 ) (1 .5 36 ) (1 .3 71 ) (1 .3 39 ) N u m b er o f si b lin g s -0 .2 91 -0 .0 32 9 -0 .0 79 1 0. 03 25 0. 93 3* * 0. 3 62 0. 21 3 -0 .1 52 (0 .6 55 ) (0 .4 4 0) (0 .2 8 3) (0 .1 86 ) (0 .3 82 ) (0 .4 26 ) (0 .3 03 ) (0 .2 51 ) S F- 12 P C S -0 .0 97 1 -0 .1 32 * -0 .2 69 ** * -0 .2 8 8* ** 0. 10 9* * 0. 06 32 -0 .0 11 7 -0 .0 87 1 (0 .1 17 ) (0 .0 72 2) (0 .0 6 65 ) (0 .0 23 8) (0 .0 55 3) (0 .0 79 9) (0 .0 55 2) (0 .0 5 40 ) M o th er ’s a g e (z -s td .) -0 .1 8 5 -0 .2 19 -0 .1 79 -0 .1 17 0. 3 30 -0 .1 3 4 0. 3 31 0. 4 93 (0 .6 17 ) (0 .4 4 8) (0 .3 21 ) (0 .1 5 8) (0 .5 15 ) (0 .5 0 8) (0 .3 77 ) (0 .3 14 ) C on st an t 52 .9 1* ** 59 .6 2* ** 71 .4 8* ** 76 .3 3* ** 35 .2 7* ** 47 .1 2* ** 55 .8 3* ** 65 .2 1* ** (6 .9 8 4) (4 .2 78 ) (3 .9 72 ) (1 .4 76 ) (3 .6 20 ) (4 .7 11 ) (3 .1 83 ) (3 .1 95 ) O b se rv at io n s 1, 16 8 1, 16 8 1, 16 8 1, 16 8 1, 16 9 1, 16 9 1, 16 9 1, 16 9 N ot e: R o bu st s ta nd ar d er ro rs in p ar en th es es . S ig ni  c an ce le ve ls : ** * p< 0. 00 1, * * p< 0. 01 , * p < 0. 05 . S ou rc e: G er m an S o ci o -E co no m ic P an el (S O E P) a nd S O EP -R V li nk • Lara Bister, Jeroen Spijker, Fanny Janssen, Tobias Vogt736 Ta b . A 3. 4: A ve ra ge m ar g in al e ff ec t ta b le s fo r S F- 12 M en ta l C o m p o ne n t S co re (M C S ) ( N = 2 ,3 37 ) M en W o m en Q u an ti le s Q u an ti le s A ve ra g e m ar g in al e ff ec ts 20 % 4 0% 6 0% 8 0% 20 % 4 0% 6 0% 8 0% B o rn in E as t G er m an y -2 .0 75 -1 .5 02 -2 .5 9 5* ** -0 .7 9 4 -1 .9 9 8* -3 .3 9 6* ** -3 .1 70 ** * -2 .9 08 ** * (r e f.: W es t G er m an y) (1 .4 3 3) (0 .9 07 ) (0 .6 8 3) (0 .5 15 ) (1 .0 14 ) (1 .0 8 3) (0 .7 55 ) (0 .6 79 ) B o rn in E as t G er m an y X M ig ra ti o n to W es t G er m an y N o m ig ra ti o n -2 .1 83 -2 .1 56 * -3 .0 70 ** * -1 .5 3 6* -1 .9 13 -3 .0 3 8* -2 .7 52 ** * -2 .8 74 ** * (1 .5 85 ) (1 .0 78 ) (0 .8 07 ) (0 .6 4 4) (1 .1 8 9) (1 .2 3 0) (0 .8 76 ) (0 .8 22 ) M ig ra ti o n -1 .8 14 0. 06 82 -1 .4 55 0. 98 5 -2 .1 94 -4 .2 29 * -4 .1 42 ** * -2 .9 8 7* * (2 .8 13 ) (1 .5 9 5) (1 .1 76 ) (0 .8 23 ) (1 .9 3 6) (1 .9 96 ) (1 .2 2 0) (1 .2 31 ) B o rn in E as t G er m an y X A n y p ar en ta l u n em p lo ym en t 19 9 0 -1 99 4 N o u n em p lo ym en t -2 .0 82 -1 .2 6 5 -2 .2 94 ** * -1 .5 57 * -2 .5 52 * -2 .3 62 -1 .8 47 * -2 .4 78 ** * (1 .6 5 4) (1 .0 16 ) (0 .7 41 ) (0 .7 3 4) (1 .2 5 0) (1 .2 4 0) (0 .7 4 6) (0 .8 5 9) U n em p lo ym en t -2 .0 62 -1 .9 06 -3 .1 0 8* 0. 50 3 -1 .1 52 -4 .9 75 * -5 .1 91 ** * -3 .5 6 4* ** (2 .6 26 ) (1 .8 37 ) (1 .4 16 ) (0 .7 26 ) (1 .8 03 ) (2 .0 59 ) (1 .6 02 ) (1 .2 0 4) O b se rv at io n s 1, 16 8 1, 16 8 1, 16 8 1, 16 8 1, 16 9 1, 16 9 1, 16 9 1, 16 9 N ot e: R o bu st s ta nd ar d er ro rs in p ar en th es es . S ig ni  c an ce le ve ls : ** * p< 0. 00 1, * * p< 0. 01 , * p < 0. 05 . S ou rc e: G er m an S o ci o -E co no m ic P an el (S O E P) a nd S O EP -R V li nk Scarred for Life • 737 Calculation of the SF-12 summary score Description The SOEP version of the SF-12 summary score for physical and mental health functioning (SF-12) has been included biannually in the SOEP since 2002 and is used as a reliable tool of meas-uring self-reported health functioning and health-related quality of life (Andersen et al. 2007). The SF-12 is a subset of the SF-36v2® that measures eight domains of health, which are used to calculate a physical (PCS) and a mental component score (MCS): role limitation due to physical problems, general health perceptions, vitality, bodily pain, physical functioning, social function, role limitations due to emotional problems, and general mental health (RAND Corporation 2020). The SF-12 is known as a valid and reliable standardised tool for assessing mental health functioning and health-related quality of life (Andersen et al. 2007; RAND Corporation 2020). • Lara Bister, Jeroen Spijker, Fanny Janssen, Tobias Vogt738 It em S O E P va ri ab le S u rv ey q u e st io n A n sw er c at eg o ri es M et h o d o f i te m c re at io n P hy si ca l p le 00 04 W h en y ou h av e to c lim b 1 S tr o ng C o m b in ed a s m ea n fu n ct io n in g (P F) se ve ra l  ig ht s o f s ta ir s, d o es 2 S lig h tl y yo u r cu rr en t h ea lt h li m it yo u? 3 N o t at a ll p le 00 05 A n d w h at a b o u t o th er d em an d in g ev er yd ay 1 S tr o ng ac ti vi ti e s, s u ch a s w h en y o u h av e to li ft 2 S lig h tl y so m et h in g h ea v y o r d o s o m et h in g 3 N o t at a ll r eq u ir in g p h ys ic al m o b ili ty : D o es y o u r h ea lt h lim it y o u? G en er al h ea lt h (G H ) p le 00 08 H o w w o u ld y o u d es cr ib e yo u r cu rr en t h ea lt h? 1 V er y go o d R ev er se d a n sw er 2 G o o d ca te go ri es f ro m 3 S at is fa ct o ry 1 = b ad t o 5 = v er y go o d 4 P o o r 5 B ad B o d ily p ai n ( B P ) p le 00 3 0 H o w o ft en in th e la st fo u r w ee ks d id y o u 1 A lw ay s N o a d ju st m en t su ff er f ro m s ev er e p h ys ic al p ai n? 2 O ft en 3 S o m et im es 4 A lm o st n ev er 5 N ev er R o le p h ys ic al ( R P ) p le 00 31 H o w o ft en in th e la st fo u r w ee ks , d u e to 1 A lw ay s C o m b in ed a s m ea n p hy si ca l p ro b le m s, d id y o u ac h ie ve le ss 2 O ft en in y o u r w o rk o r ev er yd ay a ct iv it ie s th an 3 S o m et im es yo u ac tu al ly in te n d e d ? 4 A lm o st n ev er 5 N ev er p le 00 32 H o w o ft en in th e la st fo u r w ee ks d id y o u 1 A lw ay s ex p er ie nc e lim it at io n s d u e to p h ys ic al 2 O ft en p ro b le m s? 3 S o m et im es 4 A lm o st n ev er 5 N ev er Ta b . A 4: S O EP s u rv ey q u es tio n s an d m et ho d o f cr ea tin g th e S F- 12 it em s fo r ca lc ul at in g th e p hy si ca l ( P C S ) a nd m en ta l co m -p o ne n ts s co re s (M C S ) f o r p hy si ca l a nd m en ta l h ea lth fu nc tio n in g Scarred for Life • 739 Ta b . A 4: C o n tin ua ti o n It em S O E P va ri ab le S u rv ey q u e st io n A n sw er c at eg o ri es M et h o d o f i te m c re at io n M en ta l h ea lt h (M H ) p le 00 27 H o w o ft en in th e la st fo u r w ee ks d id y o u 1 A lw ay s C o m b in ed a s m ea n fe el in lo w s p ir it s an d m el an ch o ly ? 2 O ft en (r ev er se d a n sw er 3 S o m et im es ca te go ri es o f se co n d 4 A lm o st n ev er q u es ti o n f ro m 5 N ev er 1 = n ev er t o 5 = a lw ay s) p le 00 28 H o w o ft en in th e la st fo u r w ee ks d id y o u 1 A lw ay s fe el w el l- b al an ce d ? 2 O ft en 3 S o m et im es 4 A lm o st n ev er 5 N ev er R o le e m o ti o n al ( R E ) p le 00 3 3 H o w o ft en in th e la st fo u r w ee ks , d u e to 1 A lw ay s C o m b in ed a s m ea n p sy ch o lo g ic al o r em o ti o na l p ro bl em s, 2 O ft en d id y o u ac h ie ve le ss in y o u r w o rk o r 3 S o m et im es ev er yd ay a ct iv it ie s th an y o u ac tu al ly 4 A lm o st n ev er in te nd ed ? 5 N ev er p le 00 3 4 H o w o ft en in th e la st fo u r w ee ks , d u e to 1 A lw ay s p sy ch o lo g ic al o r em o ti o na l p ro bl em s, 2 O ft en w er e yo u le ss c ar ef u l i n y o u r w o rk o r 3 S o m et im es ev er yd ay a ct iv it ie s th an y o u ac tu al ly 4 A lm o st n ev er in te nd ed ? 5 N ev er S o ci al p le 00 3 5 H o w o ft en in th e la st fo u r w ee ks , 1 A lw ay s N o a d ju st m en t fu n ct io n in g (S F) d u e to h ea lt h o r ps yc h o lo gi ca l p ro b le m s, 2 O ft en h av e yo u b ee n re st ri ct ed in t er m s o f 3 S o m et im es yo u r so ci al c o n ta ct t o f o r ex am p le f ri en d s, 4 A lm o st n ev er ac q u ai n ta n ce s o r re la ti ve s? 5 N ev er V it al it y (V T ) p le 00 29 H o w o ft en in th e la st fo u r w ee ks d id y o u 1 A lw ay s R ev er se d a n sw er fe el f u ll o f en er gy ? 2 O ft en ca te go ri es f ro m 3 S o m et im es 1 = n ev er t o 5 = a lw ay s 4 A lm o st n ev er g o o d 5 N ev er S ou rc e: S O EP C o m pa ni on T op ic s of H ea lth a n d C ar e [h tt p: // co m pa ni on .s oe p. d e/ To pi cs % 20 of % 20 S O EP co re /H ea lth % 20 an d% 20 C ar e. ht m l] • Lara Bister, Jeroen Spijker, Fanny Janssen, Tobias Vogt740 References Andersen, Hanfried H. et al. 2007: Computation of Standard Values for Physical and Mental Health Scale Scores Using the SOEP Version of SF12v2. In: Ferrer-i-Carbonell, Ada; Grabka, Markus M.; Kroh, Martin (Eds.): Schmollers Jahrbuch: Journal of Applied Social Science Studies 127,1: 171-182. RAND Corporation 2020: 12-Item Short Form Survey (SF-12) [https://www.rand.org/ health-care/surveys_tools/mos/12-item-short-form.html, 27.10.2020]. Published by Federal Institute for Population Research (BiB) 65180 Wiesbaden / Germany Managing Publisher Dr. Nikola Sander 2023 Editor Prof. Frans Willekens Managing Editor Dr. Katrin Schiefer Editorial Assistant Beatriz Feiler-Fuchs Wiebke Hamann Layout Beatriz Feiler-Fuchs E-mail: cpos@bib.bund.de Scientifi c Advisory Board Kieron Barclay (Stockholm) Karsten Hank (Cologne) Ridhi Kashyap (Oxford) Natalie Nitsche (Rostock) Alyson van Raalte (Rostock) Pia S. Schober (Tübingen) Rainer Wehrhahn (Kiel) Comparative Population Studies www.comparativepopulationstudies.de ISSN: 1869-8980 (Print) – 1869-8999 (Internet) Board of Reviewers Bruno Arpino (Barcelona) Laura Bernardi (Lausanne) Gabriele Doblhammer (Rostock) Anette Eva Fasang (Berlin) Michael Feldhaus (Oldenburg) Alexia Fürnkranz-Prskawetz (Vienna) Birgit Glorius (Chemnitz) Fanny Janssen (Groningen) Frank Kalter (Mannheim) Stefanie Kley (Hamburg) Bernhard Köppen (Koblenz) Anne-Kristin Kuhnt (Rostock) Hill Kulu (St Andrews) Nadja Milewski (Wiesbaden) Roland Rau (Rostock) Thorsten Schneider (Leipzig) Tomas Sobotka (Vienna) Jeroen J. A. Spijker (Barcelona) Heike Trappe (Rostock) Helga de Valk (The Hague) Sergi Vidal (Barcelona) Michael Wagner (Cologne)