SOEP Conferences

4th International German Socio-Economic Panel User Conference (SOEP2000)

GSOEP2000: Abstracts

July 5, Second Keynote

Family Change, Employment Transitions, and the Welfare State: Household Income Dynamics in the United States and Germany

Thomas A. DiPrete, Duke University
Patricia A. McManus, Indiana University

Since the demise of modernization theory, social scientists have sought explanations for persisting differences in the stratification of industrialized societies, primarily by studying how educational and labor market institutions shape the life chances of individuals. This approach undervalues two key features of any stratification system: family dynamics and the welfare state. Employment changes, changes in household composition, and changes in the employment situation of a spouse or partner can all trigger large shifts in income and material well-being. The impact of these events is mediated by public tax and transfer mechanisms and by private actions taken by household members. This comparative analysis of household income dynamics in the United States and Germany shows that variations in welfare state policy produce distinct societal patterns of income mobility, and furthermore shows that the relative importance of labor market events, family change, and welfare state policies for income dynamics depends on gender. The strong interrelationship between individual incentives and the structure of opportunity produces an asymmetry in the long-term impact of events. The negative effects of events that reduce income generally decay over time, while the effects of positive events generally persist.

nach oben

July 6 a.m., Session A: Structure of Wages

Pathbreakers and Traditionalists - Parental Resources, Career Credentials and Self-Employment Outcomes in the U.S. and Germany

Patricia McManus
, Indiana University

Using longitudinal data from the Panel Study of Income Dynamics and the German Socioeconomic Panel, this research compares pathways into self-employment among men and women in the United States and Western Germany. Higher general academic credentials are associated with self-employment entry and stability in the U.S., while specific occupational credentials in Germany facilitate entry into self-employment but do not provide a significant source of self-employment stability. Intergenerational transmission of self-employment is more among men than among women in both countries, while spousal transmission of self-employment status is more prominent among women. Career credentials are an important source of earnings stratification in the self-employment sector in both countries, but parental self-employment has little impact on one's own self-employment earnings. In both countries, women's self-employment mobility is sensitive to domestic responsibilities.

Early Career Experiences and Later Career Outcomes - A Comparison of the United States, France and Germany

David N. Margolis, CNRS/ University of Paris I
Véronique Simonnet, University of Paris I
Lars Vilhuber, U.S. Census Bureau/ York University Toronto

This paper represents an attempt to untangle the theoretically complicated, and empirically uncertain, links between the early career experiences of young people in the labor market and their labor market success or failure later in life. We use a large set of measures of early career experiences and consider the impact of these measures on a variety of different measures of later career success 5 to 20 years after labor market insertion. Institutions seem to condition labor market outcomes, as our results differ significantly from one country to the next, for the same variables in the same econometric specifications. We find that labor market outcomes in (western) Germany are consistent with a dual labor market model. In the case of American workers, either the market learns about unobservable worker characteristics over time, or the implicit contracts established at the start of the career are increasingly renegotiated over time. Unobserved heterogeneity in individuals' networks of labor market contacts is an explanation that seems to be the most consistent with our results for France. These results reflect optimal firm responses to the different institutional environments in each country in the presence of imperfect information concerning young workers ex ante.

Individual Heterogeneity, Job Matching, and Returns to Tenure in Germany

Kenneth A. Couch, University of Connecticut

Data from the German Socio-Economic Panel are used to examine the roles of individual heterogeneity and job match quality in generating commonly observed wage-tenure profiles. The evidence presented in the paper indicates that once those factors are reflected in the estimations, the returns to seniority are no longer measurable. Job math quality appears to be the dominant factor in the German labor market in generating an upward sloping wage-tenure profile.

What Determines the Reservation Wage? Theory and Some New Evidence from German Micro Data

Eswar S. Prasad, IMF

This paper constructs a simple theoretical model that relates the reservation wage of unemployed workers to macroeconomic factors--including aggregate and local unemployment rates, generosity of the unemployment compensation system and characteristics of the wage structure--as well as individual-specific determinants, including proxies for general and specific human capital, length of unemployment spell and alternative income sources. The predictions of the model are tested using data from the German Socio-Economic Panel.

The longitudinal aspect of the dataset provides an interesting perspective on how reservation wages change over time and how they correlate with accepted wage offers for workers who make the transition from unemployment to employment. The findings have important policy implications as well, since they shed some light on the disincentive effects of the German tax and transfer system for the labor supply and employment decisions of unemployed workers at different points of the skill/offer wage distribution.

nach oben

July 6 a.m., Session B: Earnings and Income Mobility

Cross-National Estimates of the Intergenerational Mobility in Earnings

Dr. Dean R. Lillard, Cornell University and DIW

This paper examines the similarity in the association between earnings of sons and fathers in Germany and the United States. It relaxes the log-linear functional form imposed in most studies of the intergenerational earnings association. Theory implies the relationship between earnings of fathers and sons could be non-linear, especially at the tails of the fathers' earnings distribution. When a more flexible function form is fit to the data, the apparent similarity between Germany and the United States disappears. Relative to mobility in Germany, upward mobility is higher in the United States for sons from the poorest fathers and downward mobility is lower for sons from fathers with high earnings.

Income Mobility in the United States and Germany - A Comparison of Two Classes of Mobility Measures using the GSOEP, PSID, and CPS

Andrew J. Houtenville, Cornell University

The United States is often considered to be more free-wheeling and mobile than Germany; however, previous cross-national studies of income mobility find the opposite is true. This paper investigates these surprising results and finds that the results of previous studies are confirmed when income mobility is measured by changes in the positions of individuals in the income distribution-members of former West German households are more income mobile than Americans. However, conventional belief is confirmed when income mobility is measured by absolute movements in income-Americans are more income mobile than members of former West German households.

Local versus global assessments of mobility

Chris Schluter, University of Bristol
Mark Trede, University of Cologne

The common approach to measuring income mobility is to compute a mobility index, which reduces the information about income changes contained in the joint distribution of incomes into a scalar. Information about "local" income changes is aggregated into a "global" mobility index. We derive an approximation to the aggregation rule for the important class of so­called stability indices. By comparing global mobility estimates and local distributional change between the USA and Germany, we explain the empirical puzzle observed by Burhauser et. al. (1997) who found that Germany has more income mobile than the USA. We show that the relative global mobility ranking of the two countries is driven by the first­period poors.

Toward a Longitudinal, Multi-Dimensional Class Model

James C. Witte, Clemson University

A longitudinal analytical framework, one that sees class as a process over time and not a fixed attribute is proposed as a means to redirect class analysis and revive a theoretical debate that has gone stale. Class analysis implies an inherently dynamic perspective. However, quantitative studies of class that go beyond static analyses of cross-sections are rare. Three dimensions of class may be identified in previous work on the effects of class: 1) class as market situation, 2) class as social status, 3) class as position in the relations of production. A longitudinal multi-dimensional model would allow each of these aspects of class to independently influence the life course over time. Hierarchical Linear Modeling techniques (HLM) are used to estimate a model of this type. Thirteen years of individual income data from the German Socio-Economic Panel (GSOEP) is used to illustrate the application of this framework. The estimate results indicate that each of the three dimensions significantly affects income trajectories over time, but they do so in different ways. Class as market situation and class as position in the relations of production generally affect the level of income, but not the change in income over time. Class, as social status, on the other hand, typically has little effect on the initial income level but does influence the patterns of earnings growth over time. While this particular pattern of results may be limited to the German case, the proposed longitudinal, multi-dimensional modeling framework is widely applicable.

nach oben

July 6 a.m., Session C: Work/Family Processes and Socioeconomic Outcomes

Lags and Leads in Life Satisfaction - A Test of the Baseline Hypothesis

Andrew E. Clark, CNRS and LEO, University of Orléans;
Ed Diener, University of Illinois at Urbana-Champaign;
Yannis Georgellis, Brunel University

This paper uses fourteen waves of the German socio-economic panel (GSOEP) data set to examine the long-run impact of life and labour market events on life satisfaction. By doing so, we advance from the standard literature which almost exclusively considers contemporaneous correlations. Our main interest is to test a revered idea in Psychology, that of the "baseline" level of life satisfaction. According to this hypothesis, individuals will, in the long run, return to some baseline level of well-being or utility, independently of what has happened to them in the past. This hypothesis is obviously similar to the economist's and psychologist's notion of habituation, except that we consider life events, rather than consumption, income, or behaviour.

We consider six different life events: unemployment, layoff, quitting a job, marriage, divorce, and birth of first child. We use the sub-sample of West Germans aged between 19 and 59. This produces approximately 90 000 person-year observations, split roughly 50:50 between men and women. Our proxy utility measure is overall life satisfaction, measured on a scale of zero to ten.

As a first step, simple graphical representation of average life satisfaction (for those affected versus a control group) at the time of the event, and over the twelve years following the event, shows that the strongest effect is often at the time that the event happens. However, significant differences in life satisfaction between the affected group and the control group persist. In the case of unemployment and layoffs, we still see significant differences six or eight years after the event. The effect of quitting a job seems to disappear after four years. Marriage, divorce, and the birth of first child are all associated with significant differences in life satisfaction, which persist for five years or more. However, all of the graphs show clear signs of adaptation.

We then carry out ordered probit regressions, controlling for age, marital status, education, number of children, household income, regional and year dummies. All of our analysis is carried out separately for men and women. We find strong evidence of lag and lead effects in life satisfaction with respect to the six job and life events described above. We also find evidence of adaptation, in that past experiences become less important. Our first results suggest, tentatively, that men adapt less quickly than women to labour market events, in that past unemployment, quits and layoffs continue to be important for men for a longer time than they do for women.

Timing of First Births in East Germany after Unification

Michaela Kreyenfeld, Max Planck Institute for Demographic Research (Rostock)

When German unification was accompanied by a rapid decline in aggregate fertility rates, researchers particularly assigned high unemployment rates a dominant role for changes in fertility behavior. The aim of this paper is twofold. First, we investigate changes in the timing of first birth in East Germany after unification. Using data from the SOEP, we show that even after unification East Germans are younger at first birth than their West German counterparts. Second, we investigate the relationship between male and female unemployment and first birth risks. A major result is that female unemployment fosters the transition into parenthood in East Germany.

The Relation between Success at Work, Life Pattern and Overall Life Satisfaction - Changes in Men´s and in Women's Lives since the 1980s

Angelika Toelke, DJI

The relationship between family building and employment history has been one of the main topics in research on the family as far as women are concerned. In contrast, research on occupational success of men rarely takes note of live events related to their family and/or to their personal relationships. Only very recently, attention has been directed towards the implications of dual careers and the importance of family for male occupational success in general. Up to the 1970s, the consensus in family economics was that a gender specific division of labor was beneficial for all family members. But as new patterns of intimate relationships have emerged, it can no longer be taken for granted that people are situated in a (traditional) family during their middle age. New assertions about the connection between intimate relationship and occupational mobility have emerged. Singles and/or people without children are seen as being more flexible in their time management, as having better chances to continue their education, and as being more mobile geographically - they are seen as the winners. However, these assumptions have not yet been put to a stringent empirical test. The paper will focus on the question whether there is an effect of life patterns on the professional career of men. What is the structural relationship between the two life domains? Does the change in life patterns have an effect on that relationship? To what extent do success at work and specific life patterns affect overall satisfaction with life? The analysis uses data for the subsample of men aged 30 to 50 from the 1986 and 1996 SOEP. Possible career advances between their first occupation and their occupation at the time of the interview are analyzed in view of other life course events using logistic regression models. Changes, from 1986 to 1996, in the structural relationship between the two domains as well as changes in the effect on overall life satisfaction are considered as important indicators of social change.

Do Immigrants Improve their Housing Quality when They Move? Evidence from the German Socio-Economic Panel

William A. V. Clark and Anita I. Drever, University of California, Los Angeles

Previous research using the German Socio-Economic Panel showed that immigrants moved slightly more frequently than native born Germans. The research in this paper extends that work and examines the extent to which this increased mobility is translated into improved housing quality. Overall, we find that all households have improved their housing status over time and that both the foreign born and German households gain additional housing space when they move. At the same time, immigrant households have yet to achieve parity in the number of rooms, square meters, or levels of satisfaction. Interestingly, even though both immigrants and Germans have increased the amount of space they consume, they are less satisfied than a decade ago. The logit models emphasize that the classic dimensions, age, tenure choice and household size explain the likelihood of increasing space with mobility. The models are much less accurate predictors of levels of satisfaction or degree of building renovation.

nach oben

July 6 p.m., Session A: Job Mobility and Unemployment

The Duration of Marginal Employment in Germany

Jürgen Kolb, Institute for Economic Research Halle
Axel Werwatz, Centre for Labour Market and Social Research, Århus

Since the 1980s there has been a big discussion in Germany regarding jobs with wages below the Lower Earnings Limit (LEL) for mandatory social security contributions. On one hand, these jobs seem to be a source of flexibility in a labour market that has frequently been characterized as being overly rigid. On the other hand, these jobs offer little (if anything) in terms of social protection.

Although cross-section studies show a rise in the number of LEL-jobs, little is known about the duration of these jobs and their role within people‘s labour market careers. This paper tries to fill this gap. Using spell data based on 15 waves of the German Socio-Economic Panel (GSOEP) we present a longitudinal analysis of LEL-jobs in Germany. After discussing some problems of identifying LEL-jobs with GSOEP-spelldata, we present results suggesting that LEL-jobs frequently cover a short period of a person‘s labour market history.

We then develop a setting to analyze the impact of socioeconomic variables on duration of LEL-jobs. We first analyze the impact of variables describing the individual holding an LEL-job (sex, age, etc.). Based on ideas from the economic theory of the family and labour supply we extend our analysis to include variables describing the individual's family/household situation (income, children, spouse, etc.). We also control for aspects of the German tax system and the regulations concerning maternity-leave. Using semiparametric Cox-regression we find rather strong evidence for the impact of age, sex, human capital and some evidence for the influence of family characteristics.

Why Do German Firms Subsidise Apprenticeship Training? Tests of the Asymmetric Information and Mobility Cost Explanations

Damon Clark, Nuffield College/LSE

It is often observed that despite the famous prediction of Becker (1962) that firms will not pay for general training, German firms do in fact subsidise Apprenticeship Training. This paper examines two prominent solutions to this puzzle - ‘asymmetric information' and ‘mobility costs'. Our tests do not support the asymmetric information hypothesis, and, whilst they provide evidence consistent with a simple mobility cost explanation, we argue that this hypothesis is deficient in a number of other respects.

German Wage Underpayment - An Investigation into Labour Market Inefficiency and Discrimination

Peter Dawson, Albertay University,
Timothy Hinks, Middlesex University,
Duncan Watson, Middlesex University

Using stochastic panel wage frontiers, this paper estimates the relative underpayment of females and males in the reunified Germany. The estimates are initially applied to discrimination analysis. It finds that females have higher underpayment and that the male-female differential is increasing over the period 1991-1993. However, the paper suggests that the underpayment estimates reflect other concerns, such as dynamic monosony. Labour market inefficiency must be taken into account before discrimination analysis is possible.

nach oben

July 6 p.m., Session B: The Structure of Inequality

(Not) Sharing the Fruits of Growth - The Impact of Governments and Markets on Living Standards in Germany, the Netherlands and the U.S.A., 1986-95

Bruce Headey, Melbourne University
Stephen Headey, Melbourne University
Ruud Muffels, Tilburg University

The 1980s and 1990s have been decades of quite good economic growth in North America and much of Western Europe. But how have the fruits of growth been distributed? This paper reviews changing income distributions in the U. S, Germany and the Netherlands. These three countries may be taken as exemplars and leading economic performers in ‘the three worlds of welfare capitalism' (Esping-Andersen, 1990). The U.S. is a liberal welfare-capitalist state, Germany a coporatist state, and the Netherlands (less clearly) a social democratic welfare-capitalist state. The paper focuses particularly on income changes in the bottom, middle and top quintiles in the decade 1986-95. Previous analyses have shown that labour earnings dispersion is increasing, with increased returns to human capital. The potential impact of government through the tax-transfer system has been largely ignored. This paper shows that, while labour income dispersion has increased in all three countries, the Dutch government alone has redistributed, with the result that the bottom quintile has gained along with others. The U.S., Germany and the Netherlands are the only three countries for which over ten consecutive years of panel data are available. The data used here come from the PSID-GSOEP Equivalent File 1980-97 and from a similar file constructed from the Dutch SEP data.

The Gender Gap During Rapid Economic Transition - Women's Economic Progress Since Transition

Ira N. Gang, Rutgers University
Myeong-Su Yun, University of Western Ontario

East Germany underwent a rapid transition from a socialist to a market economy. We are interested in whether women are better off or worse off in both absolute terms and relative to men as a result of this transition. We use the German Socio-Economic Panel Data 1990-1997 to jointly study wages, hours of work and participation in the labor market. We implement a decomposition analysis which accounts correctly for selection bias issues and jointly considers the determinants of wages, hours of work and participation in the labor market. In this way we analyze the complete labor market effects of rapid transition on former East German citizens since the German unification.

Gender Effects As Macro-Level Effects

Lisa M. Amoroso, Northwestern University

This paper illuminates a unique approach to modeling structural effects on individual outcomes; more specifically, gender effects as macro-level effects on individual labor market compensation. Feminist sociology has demonstrated that gender is far more than an individual characteristic, but is better theorized as an institutional regime or organizing principle constraining and enabling the activities of groups of individuals in a disproportional ways. Gender is more than an individual characteristic and can be modeled as a group membership effect using quantitative modeling techniques.

This paper examines the effect of group memberships on annual income among U.S. and Germany residents from 1984 to 1996. I consider the institutional effects of demographic groups such as gender, race/citizenship, marital status, family composition, and country. The central thesis of this paper is that group level characteristics inform the way in which a person operates in the labor market. This paper controls for individual characteristics such as employment status (not working, part-time, full-time) while including group level measures such as percentage within group working part-time. These data are aggregated into thirty-two demographic groups to generate variables such as the percentage breakdown of employment and education levels, as well as, average working hours. The group level measures are generated annually using cross-sectional weighted data from the German Socio-Economic Panel (GSOEP), the Panel Study of Income Dynamics (PSID), and the Syracuse PSID-GSOEP Equivalent file. I conduct regression analyses predicting the natural log of household income using robust standard errors to correct for autocorrelation. Clear and systematic group level effects exist in the US and German populations and are discussed in the paper.

The dominant ways of doing quantitative analysis do not adequately consider the contextual relationship between institutional structures and individual outcomes. Using disaggregated macro-level measures is an original contribution to comparative social research. It is a first step in addressing the complexities (specifically, of race and gender) within national borders highlighted by small case historical cross-national studies. This methodology allows for the integration of descriptive historical and contextual evidence into our casual quantitative models. In this way, this paper provides a contribution to both our theory about gender effects in the labor market and the methodology of welfare states.

Overtime Work, Overtime Compensation and the Distribution of Wages - Evidence for Germany and Great Britain

Markus Pannenberg, DIW and IZA
Gert G. Wagner, DIW/EUV/CEPS

Restrictions on overtime work, like an increase of the overtime premium or a reduction of the maximum amount of overtime hours per week, are very popular in the ongoing European debate concerning policy tools for reducing unemployment. Surprisingly, the effects of such policy tools on the distribution of wages are analyzed neither in the public nor the academic debate. Moreover, studies of the distribution of wages and their changes over time focus almost exclusively on paid-for work hours though there is evidence at least for Germany and Great Britain that unpaid overtime work is substantial.

The goal of the paper is to shed light on the impact of reducing overtime work and changing overtime compensation over the course of time on the distribution of wages and their evolution over time. Thereby we use the GSOEP for Germany and the BHPS for Great Britain, both of which provide detailed longitudinal information on overtime work and overtime compensation. In a first step we describe the incidence of overtime and the development of overtime compensation in the 1990s in both countries. For Germany, for example, this exercise shows that the incidence and the amount of overtime varies significantly with qualification, i.e. skilled workers are more likely to work overtime and exhibit higher amounts of overtime hours. Moreover, we observe a strong trend towards unpaid overtime work or compensation by leisure in Germany, in particular for the high-skilled. This points to the relevance of the analysis of the effects of overtime work on the distribution of real wages. Therefore, in a second step we analyze the effects of paid overtime work on the overall distribution of hourly real wages in Germany and Great Britain by applying a semiparametric approach suggested by DiNardo/Fortin/Lemieux (1996) to the problem at hand. This allows us to describe exactly where in the distribution of real wages paid overtime exhibit the greatest impact. In West Germany, for example, remarkable effects of overtime work are observed for 1997 only in the upper part of the hourly real wage distribution. In a third step we analyze whether the observed change in overtime compensation over the course of time, e.g. the increase of leisure compensation or unpaid work, has an impact on the evolution of wage inequality. Once again the above mentioned semi-parametric approach is applied to the problem at hand. This provides some evidence on the relative importance of changes in overtime compensation compared to changing institutions like diminishing union power, for example. In a last step we explicitly take into account the panel structure of the data. Analyzing individual overtime compensation patterns over time provide further insights into the determinants of income mobility.

The Income Portfolio of Immigrants in Germany - Effects of Ethnic Origin and Assimilation

Felix Büchel, Max Planck Institute for Human Development MPIB, and Technical University of Berlin, Dept. of Economics
Joachim R. Frick, DIW

This paper deals with the economic performance of various population groups in Germany giving special attention to ethnic origin of immigrants as compared to the native born German population. In addition, winners and losers of the re-distribution process, induced by the tax and social contributions system, are identified. This is done by considering different components of market and non-market income as well as taxes and social security contributions.

This income portfolio analysis is based on data from the German Socio-Economic Panel covering the observation years 1995 to 1997. Population subgroups are made up of West Germans, East Germans, and immigrants; the latter being split into Ethnic Migrants (Aussiedler) and foreigners. Immigrants are differentiated according to their state of assimilation: those immigrants living together with a native born adult German ("mixed" immigrants) and others ("pure" immigrants). In order to control for any retirement related impact we run all analyses for the total population as well as for those who live in households with a head of prime age.

Our results show that immigrants are a very heterogeneous group with respect to their economic performance. The well-assimilated "mixed" immigrants have an even better relative income position than the autochthonous West German population.

In general, we confirm the relatively poor market performance of immigrants. However, also immigrants´ non-market incomes are unusually low (except for those of "pure" Aussiedler), mainly due to their low eligibility for old pensions. Among younger immigrants, non-market incomes are somewhat higher than those of the autochthonous West German population, but still much lower when compared to the East German population.

Analyzing the re-distribution effect caused by public transfers, old age pensions, and tax and social security contributions, we find that immigrants as a whole are (slightly) net payers. Again, "pure" Aussiedler are an exception. When focusing only on younger people, the situation is reversed, and immigrants profit slightly. This is due to above average non-market income of "pure" immigrants. However, East Germans are still much better off.

Using random-effects models to control for various socio-economic measures, we find that non-German immigrants remain net payers. In addition, the risk of burdening the re-distribution process rapidly and strongly decreases with progressing duration of stay in Germany. On the other hand, the subgroups substantially profiting from the re-distribution process are "pure" Aussiedler and East Germans. This means --at least in our static "snapshot" analysis-- that the "classic" immigrants (of non-German nationality) are no economic burden to the autochthonous population.

nach oben

July 6 p.m., Session C: Health, Aging and Retirement

Long-Term Labor Force Exit and Economic Well-Being - A Cross-National Comparison of Public and Private Income Support

Richard V. Burkhauser, Cornell University
Dr. Dean R. Lillard, Cornell University
Paola M.Valenti, Cornell University

Public policy makers have long been concerned with the economic well-being of individuals and families as one or more family member permanently exits the labor force. Long-term exit from the labor force can occur at differing ages for differing reasons. Individuals retire or become disabled or unemployed for long periods of time. This paper examines and compares economic well-being in a sample of individuals who are observed to exit from the labor force into a longer period of no labor force participation. The paper focuses on documenting and understanding the economic well-being of these individuals and their families just before and just after they exit the labor force. The paper uses data from the US (Panel Study of Income Dynamics-PSID), Germany (German Socio-Economic Panel-GSOEP) and Britain (British Household Panel Study-BHPS). Although we measure individual incomes, the main focus of the paper is on each individual's share of household income.

Included in the measures of income are income from labor earnings, income from assets including dividends, interest payments and stock dividends, imputed rental income, income from public pensions, income from private pensions, income from public transfer programs and income from persons outside the household.

The main research questions to be considered are:

* How well does the social safety net in each country protect workers and their families from substantial drops in economic well-being following their labor force exit?
* What sources of income do individuals have just before and just after labor force exit?
* How do those sources vary before and after they exit the labor force?
* How important is income of other household members?
* How is the mode of labor force exit correlated with the rate at which income is replaced?
* What similarities and differences do we observe between the US, Germany and Britain?

The PSID, GSOEP and BHPS are longitudinal data that provide detailed information on the incomes of a random sample of men and women in each country. The PSID data are drawn from a panel of 5,000 households first surveyed in 1968. All individuals in the original households and all individuals joining their households have been sampled. The GSOEP data are drawn from an ongoing longitudinal study that began in 1984 with a survey of roughly 12,290 individuals in 5,921 German households. All individuals from these original households have been resurveyed fourteen times as well as new sample members who join the household of one of the original survey respondents. Several new samples, of immigrants and former East German citizens have been added to this core group. We focus, for institutional reasons, on the West German sample only. The BHPS panel includes information on all household members in a panel that included 10,264 individuals in 5,511 households first surveyed in 1990 and resurveyed every year since.

The study will provide insights into economic well-being of individuals as they exit the labor force. The longitudinal data used here allows us to better understand how individual well-being evolves in a more dynamic way than has been traditionally studied in cross-sectional comparisons of economic well being. The comparison of well-being across three countries, each with quite different income support schemes, also provides insights not easily drawn from a study using data from a single country.

The Dynamics of Long-Term Care Service Use in Germany

Christine L. Himes, Syracuse University,
Ulrike Schneider, Universiy of Hannover,
Douglas A. Wolf, Syracuse University

Population aging and changing family patterns has made elder care an important issue. In 1994, German lawmakers enacted a major reform in the country's long-term care policy, the Dependency Insurance Act (DIA). How, and in what way, will the relative use of formal and informal long-term care services change in response? We address this question using longitudinal data from the GSOEP to examine the mix of care providers used by older Germans prior to enactment. We find that formal care is more likely to be used by those in the poorest health, the single, or the childless. The presence of daughters increases both the use of family and formal care sources. Future work with more recent waves of the GSOEP is needed to see if family care provision is sustained in an environment of universal public long-term care insurance.

The Effect of Job Displacement on Subsequent Health

William T. Gallo, Yale University School of Medicine
Elizabeth H. Bradley, Yale University School of Medicine
Stanislav V. Kasl, Yale University School of Medicine

Using data from the 1994-1996 waves of the German Socio-Economic Panel (GSOEP), this prospective, longitudinal study investigated the association between job displacement and subsequent self-assessed health (SAH). The sample studied 253 displaced workers and a comparison group of 6934 continuously-employed workers. Controlling for baseline SAH and standard demographic characteristics, we find no statistical association between job displacement and subsequent SAH. Our findings are consistent with those of earlier studies of the relationship between unemployment and subsequent health.

Economic Status of Older German Immigrants

Robert L. Clark, North Carolina State University
Anne York, Meredith College

The income of married couples in which the husband is an immigrant aged 50 and older is compared to that of native-born Germans for the period 1995 to 1997. Immigrants are divided into households that arrived in Germany before and after 1984. Using Samples A, B, and D of the GSOEP, the income of immigrants is shown to be significantly lower than that of the native German population for households age 50 to 59 and for those aged 60 and older. Differences in income appear to be due to lower earnings for the younger households and lower retirement benefits for the older households.

nach oben

July 7: Family Processes and Socioeconomic Outcomes

Income shocks, Intergenerational Transfers and Human Capital in Germany

Enrica Croda, Department of Economics, UCLA

German reunification provides a unique natural experiment in economics: the integration of a formerly centrally planned economy into a market economy. The impact of reunification on the lifetime incomes of the population has been different across groups. In the Eastern regions, retirees have benefited from an increase in pension benefits and the younger generations have seen their incomes rise sharply because of reunification-induced policies. In contrast, the financial burden of the funding of subsidies to the East has fallen mainly on West Germans. If these changes result in differing burdens across generations, families can choose to reallocate their resources to mitigate the effect of transition.

The incentives to engage in further education or training have also changed as a consequence of reunification, both for Easterners and Westerners. After reunification, Easterners have a greater incentive to acquire market-specific human capital in order to adapt to the new market-oriented environment than they did in the previous regime. Westerners may also have an incentive to update their skills because of increased competition in the labor market, but the effect for them should be smaller than the sharp change predicted for their Eastern counterparts. These incentives are stronger for the younger cohorts than for the older cohorts in both regions because the younger cohorts have a longer period of time over which to reap the returns from their investment. However, the younger cohorts are also more likely to be liquidity constrained. Financial transfers from parents could help to alleviate these liquidity constraints.

In this paper, I develop a model with altruistic parents and liquidity constrained children, and use the GSOEP data set for the period 1989-1993 to analyze the impact of the reunification shock on transfers from parents to their adult children. In the empirical analysis, 1 find that, across time, East Germans are less likely to give than West Germans. Over time, the propensity to give to all children, overall, has decreased in the East, and the amount given has increased in the West. However, transfers towards school age children have increased in both regions, even thought in the East, these children are expected to be wealthier than their parents in a lifetime sense. These results are consistent with the theoretical model proposed and support a relevant role for liquidity constraints and human capital investments in the determination of parental transfers.

The dynamics of child poverty - Britain and Germany compared

Stephen P. Jenkins, ISER, University of Essex, and DIW
Chris Schluter, Bristol University
Gert G. Wagner, DIW, European University Viadrina, and CEPS/INSTEAD

In Britain as well as in Germany throughout the 1980s and 1990s, the poverty rate among children in each year was higher than the poverty rate for adults (and the population as a whole). The only difference was that in Germany the differential was and is much smaller. From this cross-sectional perspective, Germany appears to do better for its children than Britain does. Can this conclusion be sustained once we take account of differences in poverty dynamics between the countries?

To address this question we require information for each country about children's movements in and out of poverty; whether for example there is high or low turnover among the population of poor children and who experiences short spells and long spells.

This paper addresses these longitudinal issues and provides a cross-national perspective on child poverty in Britain and Germany. The perspectives on dynamics are the particular contribution: most research for these countries has taken a cross-sectional perspective. We use comparable data from the British Household Panel Survey (BHPS) and the German Socio-Economic Panel survey. First, we provide descriptive results. In both countries a considerable number of children move in and out of poverty between one year and the next. A corollary is that the proportion of children who experience poverty at least once over a multi-year observation period substantially exceeds the cross-sectional poverty rates in any given year. Poverty incidence is clearly associated with characteristics such as family type, parental age and education, and in an expected manner. However, German children tend to move out of poverty quickly (and quicker than British children) and in the main do not experience recurrence. The risk of child poverty is thus spread over a relatively large proportion of the population, and in part reflects factors associated with the early stages of the parental life-cycle (common to all parents). Second, we investigate the correlates of entry into and exit from child poverty. We examine poverty-related events pertaining to household formation, the labour market, and the extent to which these are correlated. As it turns out, events associated with poverty entries are typically not the mirror image of the events associated with poverty exits.

Educational opportunities of children in poverty

Anke Schöb, Center for Survey Research and Methodology (ZUMA)

This paper addresses one of the central questions of the poverty research debate on direct consequences of economic deprivation for the family and the growing up of children: Are children restricted or promoted in their developmental outcomes by the economic deprivation of their families? And if so, to what extent and under which circumstances are restrictions salient? Recently, the importance of these questions has been underlined by the thesis of so-called infantilisation ("Infantilisierung") of poverty in Germany which states that an increasing share of children and young persons are located in low-income households. At the same time, empirical data shows a great diversity in the time dimension of poverty with much poverty being short-term rather than a lifelong situation. These results suggest to investigate timing and duration of poverty in childhood and raise questions about the sensitivity of developmental outcomes to both aspects of time.

In this respect two hypothesis are that, first, being poor in the early childhood is less significant for developmental outcomes than later experiences of poverty and, second, the duration of poverty influences developmental outcomes, where short periods are less disadvantageous to children than long periods of poverty. The paper addresses this subject and presents empirical results on developmental chances of children growing up in poor families. I consider educational decisions as an essential indicator for future opportunities regarding socio-economic status and income. Therefore, the choice of a secondary school (Hauptschule, Realschule, Gymnasium) that parents together with their children encounter at the end of primary school (Grundschule) is analysed. Secondly, to consider the time dimension, I take into account the (non-) poverty situations of families within the three years preceding the educational decision to be analysed. Data of analysis is taken from the German Socio-Economic Panel and based on eleven years longitudinal information obtained from individual type (1984-1995). The main sample consists of 1130 children aged between 9 and 12 years.

nach oben