Home Sociology  &  Culture  Journal Issue Article


2002/2 (Vol. 57)

  • Pages : 200
  • DOI : 10.3917/popu.202.0359
  • Publisher : I.N.E.D

Email alert

Your alert request has been correctly taken into account.
You will receive an email when new issues are put online.


Previous Pages 331 - 357 Next

German unification in 1990 provided the conditions for a “laboratory experiment” seldom possible in the social sciences. Two societies living under very different, indeed opposing, social rules, found themselves, almost overnight, placed under a single regime, that of the West. How was behaviour going to adapt? Many will remember, for example, that the prospect of relinquishing East Germany’s more liberal abortion laws provoked serious discontent in GDR, to the point of actually delaying the final agreement on reunification. In this article Dirk Konietzka and Michaela Kreyenfeld focus on non-marital births. In 1989, the proportions in east and west were 34% and 10% respectively. Would the new unified context for the economy and for family policy cause these two rates to converge? In the event, the opposite happened—the difference between them increased still further (50% in the east and 18% in the west in 1999). The authors explore the reasons for this pattern of change, and single out the sharp—and persistent—contrast in attitudes between women in the two regions as regards their attachment to the labour market and how it can be reconciled with family life.


Compared with other European countries, the former West Germany displays relatively low rates of non-marital childbearing. Since the 1960s, there has been an increase in age at first marriage, a postponement of first birth, and an increase in childlessness (Council of Europe, 2000; Dorbritz, 2000, p. 257). Childbirth and marriage remain nevertheless strongly coupled, and this has prompted researchers to speak of a child-oriented marriage in west Germany (Nave-Herz, 1994, p. 9). In the former East Germany, non-marital childbearing was relatively high by European standards, and particularly compared with West Germany. Since the 1970s, non-marital birth rates have risen steadily, reaching 33% of all births in 1989. Most researchers attribute this development to GDR family policies (Trappe, 1995, p. 210; Cromm, 1998). The intention of the government was to support single mothers, but the measures may have also encouraged women not to get married. With the end of communism and the replacement of East Germany’s institutions by those of West Germany, it was expected that the former’s inhabitants would adopt western demographic patterns, i.e. non-marital birth rates would soon fall to west German levels (Höhn and Dorbritz, 1995, p. 171; Witte and Wagner, 1995, p. 395). Contrary to this expectation, however, non-marital birth rates skyrocketed after unification, exceeding 50% of all births in 2000 (see Figure 1).

Non-marital births in East and West Germany (as percentage of all births)

Notes: Data from 2000 are preliminary estimates provided by the Statistisches Bundesamt.

Source: Statistisches Bundesamt (2001a; 2001b)

The steady increase in non-marital birth rates in the former East Germany raises several questions. In this article we focus on the role of family policies and women’s employment in explaining the unexpected increase in childbearing outside marriage. Concerning family policies, the crucial question is why non-marital births have increased even though the incentive structure in contemporary Germany is designed to give strong support to marital childbearing (Huinink, 1998, p. 301). Is this an irrational response of women or couples in the old East Germany to the new family polices? Do east Germans use the new incentive structure strategically, in a way that west Germans have failed to do for decades? Or are east Germans simply turning their backs on “traditional family forms”? Another question closely related to this issue is: do women’s employment, economic independence and work orientation weaken the role of marriage as an institution for raising children?


To investigate these issues, it is necessary to distinguish between the different “types” of non-marital births. Taking into account the modernization of family forms and new living arrangements such as cohabitation (e.g. Seltzer, 2000; Smock, 2000; Raley, 2001), we can distinguish between births to single mothers, births in marital unions, and births in cohabiting unions. Furthermore, taking a longitudinal perspective, the cohabiting couple may get married shortly after the birth of their first (or second) child, or remain unmarried permanently. Researchers initially classified cohabiting unions as “trial marriages” (e.g. Bennett et al., 1988), but it is now recognized that the cohabiting couple increasingly represents a distinct and durable family form. In this study, our primary focus is on women with children, living in stable cohabiting unions.


The remainder of this article is structured as follows. In Part I, we give an overview of the family policies that defined the relevant conditions for non-marital childbearing in East Germany before and after unification. This is the context for outlining our main hypothesis that non-marital parenthood is related to a strong work orientation among east German women. Part II contains the description of the data and the methods of analysis. In Part III, we investigate how the employment characteristics of a woman and her male partner are related to marriage decisions in eastern and western Germany. Part IV concludes.

I - Family policies and non-marital birthsin east and west Germany

1 - Family policies before unification


Family policies in the GDR were overtly pronatalist and included various provisions designed to encourage early marriage. At marriage, couples received a “home furnishing loan” of 7,000 Marks (5,000 Marks until 1986) and were given priority in obtaining their own flat. But there were also several important provisions that encouraged single parenthood. Children of single mothers received priority access to public day care. When a child was sick or when day care was unavailable, its mother was eligible for paid leave (Gysi and Speigner, 1983; Obertreis, 1986). The most important policy measure, however, was the Babyjahr introduced in 1976, under which, after the birth of her child, a single mother could take one year’s paid leave [1]  From 1961, women were allowed to take one year’s unpaid...[1]. Married mothers, on the other hand, became eligible for the Babyjahr only after the birth of a second or higher order child. Since married and unmarried couples were treated alike after the birth of the second child, this event was often the occasion for getting married (Huinink and Wagner, 1995; Huinink, 1999, p. 127).


The Babyjahr has been regarded as the main reason for the rapid increase in non-marital births (to roughly 30% of all births in the 1980s) (Höhn, 1992, p. 9). Another important factor favouring this development appears to have been changes in the housing market. In the 1970s, marriage was still an important means of obtaining a flat in the highly regulated East German housing market, but by the 1980s the housing shortage was easing somewhat and unmarried couples had less difficulty being allocated an apartment by the local authorities.


The increase in non-marital birth rates was widely interpreted as an unintended outcome of East Germany’s family policies (Trappe, 1995, p. 210). In 1986, the GDR government responded to the changes in childbearing patterns by allowing married women to take a year’s paid leave after the birth of their first child. The extension of the Babyjahr put a halt to further increase in non-marital births, though until the demise of the GDR non-marital birth rates showed no significant decline (see Figure 1).

2 - Family policies after unification


In October 1990, the two German states were united and the former East Germany’s legal and political system was in effect replaced by that of West Germany [2]  At German unification, the Unification Treaty (Einigungsvertrag)...[2]. As had been the case in East Germany, single mothers in the FRG are subject to special treatment. They are eligible for paid leave when a child is sick, and single parent status is a key characteristic increasing the likelihood of obtaining a slot in public day care facilities (Dorbritz, 1997, p. 243) [3]  Since 1992, married women (and in principle also married...[3]. A range of means-tested transfer payments (such as paid maternity leave, social welfare, and housing benefits) is also available. Single parents who are not employed enjoy priority access to these benefits since they are not living with a partner whose income would be taken into account for assessment purposes.


Although these aspects of family policy should be kept in mind during the subsequent analysis, it is unlikely that they explain the huge east-west differences in marital patterns. This is primarily because, unlike in the GDR, most policy rules in present-day Germany make an explicit distinction between cohabitation (Nichteheliche Lebensgemeinschaft), marriage and the single state (Peuckert, 1999; Schneider and Matthias-Bleck, 1999). Childrearing benefits, social welfare and housing benefits are means tested and the income of an unmarried (but cohabiting) partner is also taken into account. A similar assessment is made with regard to access to children’s day care. While children of single mothers have priority in the attribution of day care slots, children of couples in cohabiting and marital unions are usually treated alike. Since the majority of unmarried mothers in what used to be East Germany are living in cohabiting unions (see below), they are unable to take advantage of these provisions. Finally, social welfare and housing benefits are relevant only to couples with very low labour market earnings or with bleak employment prospects, or both. Despite relatively unfavourable labour market conditions, only a small proportion of men in eastern Germany are permanently out of work (Brinkmann and Wiedemann, 1995, p. 330; Mayer et al., 1999). In summary, the German tax and transfer system provides some incentives for non-working single mothers not to move in with their partners. It does not, however, discourage couples who are already cohabiting from getting married.


On the contrary, there are several important transfer payments that strongly encourage marriage. Central to this is the system of income splitting that allows married couples to file their taxes jointly. This means that the man and the woman’s incomes are added together, divided by two and taxed as separate incomes. Because of Germany’s progressive tax regime, this rule results in substantial tax relief for couples where the man and the woman have very unequal incomes. Put another way, this measure provides couples with an incentive to get married, particularly when one of the partners is permanently not working (or employed part-time) and the other is working full-time. Similar rules apply in the German health-care and pensions systems. Married housewives (and in principle “househusbands”) are covered by their spouse’s statutory health insurance and are eligible for a widow’s (or widower’s) pension.


In sum, the German institutional framework provides strong incentives for couples to get married, particularly when one of the partners withdraws from full-time employment after childbirth (e.g. Sainsbury, 1997; Drobnic, 2000). These rules have a less decisive effect, however, on east German women, since they are much more likely to be in full-time employment. In the next section, we discuss this issue in greater detail.

3 - Female employment in east and west Germany


In the early 1990s, women in what had been East Germany faced particularly unfavourable labour market constraints in the form of high female unemployment, low re-employment rates and high risks of downward status mobility once unemployed (Mayer et al., 1999; Beckmann and Engelbrech, 1999, p. 206). Some researchers predicted that most of these women would rapidly be discouraged from labour force participation and, like their counterparts in the former West Germany, would adopt the traditional “male breadwinner model” once they had a child (Dorbritz, 1997, p. 243; Huinink, 1999, p. 129).


Despite macroeconomic conditions that were consistently less advantageous throughout the 1990s, full-time employment rates among east German women remained well above west German levels. Various empirical studies have shown that east German women taking parental leave reenter the labour force sooner than their west German counterparts; while those who are currently unemployed are more actively looking for employment and are more certain that they want to work again. In addition, east German women who are in part-time employment would often prefer to extend their working hours (Engelbrech, 1997; Holst and Schupp, 1999).


East-west differences in employment patterns are particularly pronounced for women with children. Figure 2 displays the labour force participation rates of mothers by the age of their youngest child. This clearly shows that women with children are more likely to be employed full-time in east than in west Germany. When the youngest child reaches primary-school age (age 6), only 10% of west German mothers are employed full-time, 29% work part-time and 59% are not working. In the east, the pattern is almost reversed. When the youngest child reaches primary-school age, 36% of all mothers are employed full-time, and only 38% are not working.

Percentage of employed women by age of the youngest child in 1997

Notes: Employed full-time, i.e. 35 and more hours per week; employed part-time, i.e. 1-35 working hours per week; Population: women of the birth cohorts 1961-1980 who had at least one child aged 0-10 at the date of interview in 1997.

Source: Mikrozensus 1997 (our estimates).

How are we to explain the east-west differences in mothers’ labour force participation? The widely held view on this question appears to be that east German women have a stronger work orientation than their more traditional counterparts in the west. As a legacy from former socialist times, they expect economic independence and a full-time employment career as a matter of course (Braun et al., 1994; Adler, 1997). Although the striking east-west contrasts in women’s work orientation are not seriously disputed, we need to relate them to differences in the constraints on women’s labour force participation. An important point, for example, is the persistence of relatively unfavourable labour market conditions for east German men, which may have put east German women under greater financial pressure to participate in the labour force. From this perspective, high female employment rates reflect primarily not women’s striving for economic independence but the necessity for both partners to contribute to a joint household income. Furthermore, east German women experience a relatively favourable situation as regards combining childrearing and employment. Contrary to a widespread belief that unification would be followed by the immediate closure of public day care facilities (e.g. Adler, 1997, p. 44; Rindfuss and Brewster, 1996, p. 273), the east experienced a relatively broad availability of public day care throughout the 1990s. In 1998, the availability ratio of public day care for children aged 0 to 3 was only 3% in the west, but 36% in the east. Coverage of full-time care for pre-school children (age 4-6) was complete in the east, as against only 19% in the west (see Table 1) [4]  One of the main reasons for public day care coverage...[4].

Table 1 –  Availability of public day care 1990, 1994, and 1998

Ultimately, it is hard to tell whether high employment rates among east German women reflect a general attitude which views full-time work as simply a “matter of course,” or whether women with children are drawn into the labour market as a result of financial pressures. Also difficult to know is whether east German mothers are employed in larger numbers because conditions are more favourable for combining childrearing and employment, or whether work-oriented women exert greater pressure on east German communities to provide day care. Whatever the reason, a strong work orientation and more favourable conditions for labour force participation can be expected to have important implications for the decision to get married at childbirth.


As discussed above, the German tax and transfer system is particularly generous for married couples who follow a gender-specific division of labour — i.e. one of the partners devotes most of his or her time to childrearing tasks while the other is employed full-time. If both partners are employed full-time, however, marriage brings few economic advantages. It seems reasonable, therefore, to assume that east-west differences in non-marital childbearing relate primarily to differences in the labour force participation of women. In short, we expect that differences in work orientation between east and west German mothers will explain the divergent patterns in non-marital childbearing. In the following section, we investigate this hypothesis.

II - Data source and method of analysis

1 - Data source


Our empirical analysis is based on data from the German micro-census of 1997 (hereafter referred to as “Mikrozensus”) [5]  The analysis was actually conducted with data from...[5]. The Mikrozensus is a 1% sample of the resident population of Germany. In the area of West Germany it has been conducted annually since 1957 (except for the years 1975, 1983 and 1984). In what had been East Germany, the first survey was conducted in 1991 (for details, see Emmerling and Riede, 1997; Schimpl-Neimanns, 1998). It covers standard demographic characteristics (such as age, nationality, and region of residence), employment status, educational attainment, etc. The main strength of the Mikrozensus is its large sample size, which makes it particularly suitable for analysis of nuptiality and fertility patterns in the eastern and western parts of the unified Germany. Most other data sets either provide too small a sample size or cover too short a time period. For example, the German Family and Fertility Survey (FFS) was conducted during spring 1992, which means that it encompasses very little of the “post-unification period.” Other recently available data sets such as the German Socio-Economic Panel (SOEP) and the Familiensurvey provide demographic information up to 2000, but include too few cases to permit a separate analysis of “non-marital births” in east and west Germany after unification (see Huinink and Konietzka, 2002).


Although the Mikrozensus 1997 provides a sufficiently large sample size and covers a relatively long time period after unification, it suffers from one major drawback if it is to be used for the analysis of demographic events. The Mikrozensus is a cross-sectional data set that contains little retrospective information. This is a particular weakness when we wish to study the “fertility history” of the respondents. It is possible, however, to reconstruct a woman’s “fertility history” from the number of children living in the household at the time of the survey. Because this strategy involves a number of problems — for example, the older the woman, the more likely it is that her children will have already moved out of the parental home — we limited our analysis to females born between 1961 and 1980, i.e. respondents who were aged between 17 and 36 at the time of the interview [6]  According to calculations based on the Familiensurvey...[6]. Assuming that few women give birth before age 19, the children of these cohorts had a maximum age of 18 at the time of interview. We included only persons living in private households and we omitted all cases where a birth occurred before age 17.

2 - Method of analysis


The primary focus of the analysis that follows is the extent to which women’s employment affects non-marital childbearing in east and west Germany. Before proceeding with the main analysis, however, we need to set out clearly our research strategy. It involves distinguishing between single, married and cohabiting women, as well as viewing marriage from a longitudinal perspective and relating it to fertility decisions. In broad outline, our empirical analysis consists of the following three stages:

  • First, we investigate how marriage and first birth are “coupled”. In this section, we explain why we focus on women with children aged 3 to 6 and why we restrict the analysis to women in marital and cohabiting unions, i.e. why single parents are omitted from the analysis.

  • Second, we investigate the hypothesis that high non-marital birth rates reflect a strong work orientation among east German women. In this section, we use several logistic regression models to estimate the probability of living in a marital union versus a cohabiting union. The key independent variables are the activity status and the educational attainment of the woman and her partner. In addition, we make a distinction between relative and absolute educational attainment. By “relative” we mean the woman’s educational attainment compared with that of her partner.

  • Third, we focus on east-west differences in non-marital childbearing. For this part of the analysis, we pool east and west Germans in a single sample and interact various independent variables with the region of residence (i.e. east or west).

III - Empirical analysis

1 - “Unmarried parent”: a stable family status?


To what extent is non-marital parenthood a stable living arrangement? The non-marital birth rate provided by the German Statistical Office (see Figure 1) is a very crude indicator of marriage patterns since it classifies births as “non-marital” even though the couple may marry shortly afterwards. It is possible that west Germans try to avoid an “illegitimate” birth, while east Germans are more willing to postpone marriage until shortly after the birth of the first child. A further limitation is that non-marital birth rates aggregate births of different orders. It could be argued that in east Germany non-marital birth rates are high for first births, but that couples get married at the birth of the second child. A possible rationale for this is that two or more children are a serious impediment to a woman’s labour force participation.


To address this issue empirically, we consider all women who had a first or second birth between 1991 and 1996. Using life table techniques, we show how marriage is concentrated around the birth of the first or second child [7]  For this analysis, we use retrospective information...[7]. As can be seen from Figure 3, roughly 80% of the west German women are married by the end of the year in which they had their first child. In the east, this is true for only roughly 45%. When the first child reaches one year of age, 51% of the east Germans and 84% of the west Germans are married. After three years, the picture is almost unchanged: 58% of the east Germans and 86% of the west Germans are married. It follows from this that the high non-marital birth rates in east Germany do not result from a postponement of marriage to the period shortly after the first birth.


Figure 3 also displays the survival curves by age of the second child. Roughly 70% of the east German women and 90% of the west German women are married by the time their second child is born. Since second births declined rapidly after unification (Sackmann, 1999; Kreyenfeld, 2001), we can conclude that the high ratio of non-marital births is partially a composition effect.

Survival curves, transition to first marriage by age of child Percentage still unmarried by age of first and second child)
Source: Mikrozensus 1997 (our estimates).

2 - Marital union versus cohabiting union


In the following analysis, we investigate the extent to which the employment characteristics of women and their partners influence marriage decisions. The survival curves (Figure 3) reveal that hardly any marriages occur after the first child reaches three years of age — i.e. the probability of marriage is high around the birth of the first child and levels off rapidly thereafter. Against this background, we can consider as “immune” from marriage women who are still unmarried when the first child reaches three years of age. This assumption substantially simplifies the analysis. Most importantly, it allows us to switch from a longitudinal to a cross-sectional perspective. In other words, we restrict the analysis to women whose first child is at least three years old and we use the family status at the date of interview [8]  Otherwise, it would have been more appropriate to employ...[8]. Since children older than age 6 were already born (or at least conceived) when the GDR still existed, we also omit such cases from the analysis — i.e. we restrict the sample to mothers whose first children were aged between 3 and 6 in 1997 [9]  We also omitted respondents for whom information on...[9].


Table 2 displays the descriptive statistics of the remaining data set (see also Appendix). As expected, there is a much lower percentage of cohabiting unions with children in west Germany: Most unmarried west German women with children aged 3-6 years are single mothers. In the east, the opposite pattern is observed: about 16% of all women with small children are single parents and 21% are living with a partner. East German women tend to have their first child at a slightly younger age, but they are less likely to have a second child (see also Sackmann, 1999; Kreyenfeld, 2000). They are on average more highly educated and a much higher proportion of them are in full-time employment. While 34% of east German mothers are employed full-time, this is the case for only 10% in the west.

Table 2 –  Descriptive Statistics (%), population: women of the cohorts 1961-1980 with a child aged 3-6, survey year: 1997

As mentioned earlier, in the Federal Republic of Germany there are no transfer payments that actually favour cohabiting unions. Some transfers, however, might provide incentives to single parenthood, especially for women with poor labour market prospects. The descriptive statistics show that single parenthood is more common among women who have neither a college nor a vocational training diploma—which, in principle, appears to support this hypothesis. Unfortunately, we know nothing about the “partnership status” of single parents — i.e. whether they are single parents because they split up from their partner, or whether they simply never moved in with a partner. Since we are unable to incorporate the partnership status into the analysis, we will omit single parents from the subsequent analysis and concentrate on women with children (aged 3-6), who were living in marital and cohabiting unions at the date of interview.


Using a logistic regression model, we estimate the probability of living in a marital versus a cohabiting union. Our main focus of interest is the extent to which women’s employment relates to marital status. We distinguish between women who are employed full-time, part-time, and those not working. Since marriage is an institution that favours the non-participation of one partner in the labour force, we expect to find a negative correlation between women’s work and living in a marital union [10]  This implies that couples anticipate the woman’s future...[10]. In addition to women’s employment, and in line with other studies, we use educational attainment as an indicator of women’s work orientation. This variable also reflects general labour market advantages, which, particularly in the east German case, are strongly correlated with educational attainment (Brinkmann and Wiedemann, 1995, p. 330; Mayer et al., 1999). We distinguish between “no diploma,” “vocational diploma,” “college degree,” and “student” at the date of interview. In addition, we include the activity status and the educational attainment of the partner. As noted earlier, marriage is particularly advantageous when couples adopt a gender-specific division of labour — i.e. the “single earner / male breadwinner model.” In these conditions, women whose partners are unable to fulfil the role of “family provider” (because of unemployment, for example) should have a lower incentive to get married. As well as the employment characteristics of the woman and her partner, we introduce into the regression the woman’s current age and an indicator variable for having a second child.


Table 3 displays the results from the estimations. For west German women, we find the expected pattern of a strong, highly significant and negative correlation between employment and marriage. Women who are not working or working part-time are substantially more likely to be married than full-time employed women. For east Germans, however, the pattern is less clear. Similar to what is observed in the west, part-time employed women are more likely to be married than full-time employed women. But women who are not working are just as likely as women working full-time to be married.

Table 3 –  Logistic regression model, dependent variable: “marital union” versus “cohabiting union”

How should we interpret this finding? As already noted, it is important to bear in mind that east and west German women with children differ in their general orientation towards employment. While the overwhelming majority (88%) of west German mothers not employed report that they do not intend to re-enter the labour force, a large majority (65%) in the east German sample are either actively seeking employment or at least intending to return to the labour force in the near future. From this data, we conclude that in general east German non-working women do not see themselves in the role of homemaker, and that they remain “workoriented” even when not actually employed. Regarding the role of women’s educational attainment in marriage, having a college degree has a positive impact on the probability of getting married. This finding runs counter to our expectation that women who are more work-oriented would be less likely to get married. Equivocal results are also obtained for the role of the partner’s employment. In line with our expectations, having an unemployed partner has a strong negative effect on the probability of being married, but we find no correlation between the partner’s educational attainment and marriage.


It has been argued so far that in Germany, the more unequal the labour market positions of the woman and of her partner, the greater the relative gains from marriage. This means that it is not the absolute but the relative labour market situation between the woman and her partner that will have an effect on marriage decisions. In what follows, we address this issue by “combining” the woman’s educational attainment with that of her partner. We do not use the activity status at the date of interview, since educational attainment appears to be a more reliable indicator of long-term employment chances. Table 4 presents the various combinations of educational attainment of the woman and her partner. We note that assortative mating is more common in eastern than in western Germany. About 83% of east German mothers live with a partner who has the same educational level, compared with only 72% in the west. This pattern partially reflects a higher educational attainment among east German women (for details, see Wirth, 2000).

Table 4 –  Educational homogamy (percentage distribution)

Table 5 presents the results of a logistic regression model for the probability of getting married versus cohabiting, with the various combinations of the educational attainment of the woman and of her partner as independent variables. Again, the results for west Germany conform quite well to our expectations. Women who are better educated than their partners are less likely to be married (compared to couples where both partners have a vocational diploma). The results for east Germany, however, are not at all consistent with our hypothesis. Women with higher educated partners have the same probability of marriage as couples where both partners have a vocational diploma. Still more surprisingly, women who are better educated than their partners have the highest probability of marriage (of all the categories). Although studies from other countries such as the United States and Sweden have reported similar findings (e.g. Oppenheimer, 1995; Duvander, 1999), this effect is nonetheless puzzling in the German institutional context, since it appears to suggest that east German women have acquired the role of family providers

Table 5 –  Logistic regression model, dependent variable: “marital union” versus “cohabiting union”; focus of model: educational homogamy

3 - East-west comparison


The results so far indicate that employment and educational attainment have different effects in east and west Germany. By estimating separate models for the two parts of the country, however, we are unable to establish whether these differences are statistically significant. Although we know that in the east highly educated women display higher marriage probabilities than other women (particularly when they have a less educated partner), we cannot state with certainty that highly educated east German women are more likely to be married than their west German counterparts. To test whether the differences are statistically significant, we pool east and west Germans in the same sample and estimate a single regression, while allowing the east and west covariates to vary flexibly. We estimate several models, focusing in the first step on women’s absolute educational attainment (Model 1a-1c), and in the second step on women’s educational attainment relative to that of their partners (Model 2a-2e).


Table 6 displays the results from the various models, which show that for almost all subcategories, west German women are more likely to be married. The only exception consists of women who have a higher educational attainment than their partner. For this subcategory, east and west Germans do not differ significantly. This result casts new light on our previous findings. Whereas until now we would have argued that highly-educated east German women (particularly when they are living with a partner with a lower educational attainment) are more likely to get married, we now know that this is only true by comparison with east German women who hold less than a college degree. Compared with their west German counterparts, however, they are substantially less likely to be married

Table 6 –  Results from various logistic regression models with changing reference categories



In this article, we have investigated the role of women’s employment in non-marital childbearing in east and west Germany during the 1990s. We began our discussion by sketching developments in former East and West Germany. While researchers established the existence of a “childoriented” marriage pattern in the west, non-marital birth rates had increased steadily in the east since the 1970s. This development has frequently been related to peculiarities of GDR family policies, which offered single mothers more favourable rules on parental leave. Although these measures were primarily designed to improve the living conditions of single mothers, at the same time they provided an incentive for couples not to get married at the birth of a child. In the late 1980s, the GDR government finally extended parental leave to all mothers, irrespective of family status. This change in the rules brought to a halt the steady increase in non-marital birth rates, though these did not start to decline until the demise of the GDR. With German unification and the replacement of East Germany’s institutions with those of West Germany, the general expectation was that the proportion of non-marital births would soon fall to west German levels. In fact, however, they continued to increase during the 1990s, exceeding 50% in 2000.


Given that the increase in non-marital birth rates in the GDR period was largely a result of special treatment for single mothers, it is tempting to relate the continuous increase in non-marital childbearing after unification to similar “misguided” family policies. Yet in the western part of the country, where women are subject to exactly the same legal and political constraints, non-marital birth rates have remained at a low level. One obvious hypothesis for resolving this conundrum is that east German women are using the new incentives structure strategically, while west Germans have, for one reason or another, “failed” to take advantage of it.


A closer examination of FRG family policies reveals the existence of various measures designed to improve the living conditions of single parents. For example, single parents have priority access to children’s day care, welfare benefits and housing subsidies. Although such rules can create incentives to remain unmarried after childbirth, we argue that this applies only in part to the rules in the Federal Republic of Germany. The most important point is that “non-marital childbearing” covers a number of heterogeneous family forms such as single parenthood and parenthood in cohabiting unions. This distinction, which is essential from an analytical point of view, is explicitly recognized by FRG family policies that distinguish between single parents, cohabiting and married couples. Single parents get priority access to certain social transfers, whereas cohabiting and married couples are usually treated alike — i.e. the partner’s income is assessed when the couple claims social benefits. This means that although couples may be deterred from moving in together, there are no transfer payments that discourage already cohabiting couples from getting married. Since the majority of unmarried east German mothers are living in cohabiting unions, special treatment of single parents is unlikely to explain the increase in non-marital birth rates in eastern Germany in the 1990s.


Against this background, a crucial question is why east German couples in cohabiting unions are more reluctant to marry after childbirth than their west German counterparts. Our main hypothesis in this context relates to women’s employment behaviour. The German tax and transfer system — i.e. the system of income splitting, the coverage of the nonworking spouse in the national health insurance and the widow’s pension systems — is particularly advantageous for married couples who adopt the gender-specific division of labour. In other words, this system provides strong incentives for marriage — given that one of the partners greatly reduces his or her commitment to paid employment after childbirth. Yet one of the most glaring differences between the eastern and western parts of Germany in the 1990s is in the labour force participation of mothers. Compared with west German women, who give up their career, or reduce their working hours or interrupt their job for a longer period after childbirth, east German women show a completely different pattern. They return to full-time work more quickly after childbirth, while those who are unemployed express a greater desire to obtain a full-time job, and those working part-time often want to extend their working hours.


The reasons for high female employment rates in the east are manifold, however. The one most often mentioned is the high work orientation that east German women have inherited from the GDR period. Probably as important are the institutional factors. On the one hand, the unfavourable male employment situation may put east German women under greater financial pressure to be employed, while on the other, they find it easier to combine childrearing and employment, thanks to the greater availability of children’s day care. Whichever explanation is more pertinent, we assumed that a higher work orientation and more favourable conditions for combining childrearing and employment have an important bearing on marriage decisions. Given that in Germany marriage is particularly advantageous for couples who adopt the traditional division of labour, our main hypothesis was that east-west differences in non-marital childbearing relate primarily to differences in women’s labour force participation patterns.


We tested this hypothesis by estimating several logistic regression models for the probability of living in a marital versus a cohabiting union (at the date of interview for women with children aged 3-6). Activity status and educational attainment were used as indicators of a woman’s work orientation. We also controlled for the partner’s characteristics and investigated the effect of educational homogamy on the probability of living in a marital rather than a cohabiting union. The logic behind this is that the benefits of marriage are greater the more unequal the partners’ activity status. Consequently, given that education is a reliable indicator of long-term employment chances, we expected heterogamous couples to have the greatest incentive to get married.


For west Germany, we found the expected pattern. Here, a relatively high female work orientation is indeed negatively correlated with being married. F or east German y, ho wev e r, no such pattern w a s observed — women’s activity status is only very weakly correlated with marital status. Furthermore, it appears that east German women with a college degree are more likely to be married than the less educated, and women with a relatively higher educational level than their partners are significantly more likely to be married than any other reference category. East German women with a college degree, however, had a relatively higher probability of marriage only when compared with less-educated east German women. Compared with their west German counterparts, they were less likely to be married. What is more, probabilities of marriage for virtually all educational categories were significantly lower in east Germany than in west Germany.


In view of our empirical findings, it could be claimed that our main hypothesis must be rejected, since we are unable to decompose the east-west differences in non-marital childbearing using any of the variables that are commonly used to indicate women’s work orientation. The observed patterns may, however, owe something to methodological as well as theoretical arguments. A possible explanation is that standard variables such as educational attainment are inadequate indicators of east German women’s work orientation. Similarly, women’s activity status observed at the date of interview may not validly represent their long-term employment plans. Besides considerations such as these, of crucial importance are the different constraints associated with combining childrearing and employment in both parts of the country. The much more abundant provision of public day care in east Germany means that the overwhelming majority of women take for granted that they will be in full-time employment when they are mothers. In other words, pursuing a career is not, in the east, a “privilege” reserved for highly educated mothers. A wide availability of public day care and a high female work orientation, in turn, strongly reduce the economic incentives for women to get married upon childbirth. In these conditions, it is understandable why marriage, as an institution for raising children, is less prevalent in east than in west Germany.

Description of the variables

Fertility history

A woman is assumed to have given birth if she is labelled as “head of the family” or “partner of the head of the family” and if she is living in the same family with a person who is labelled as “child in the family.” The age and order of the birth is inferred from the age and the number of children in the family, i.e. the difference between the year of birth of the mother and the age of the child.

Family form


In the Mikrozensus all household members are surveyed. Apart from marital status, respondents are requested to report their relationship to the “head of the household” (i.e. whether he/she is the “partner”). Based on the marital status and information on the relationship to the “head of the household,” we distinguish the following four family forms:

  • Married women include all women who report that they are married. This is irrespective of whether they are living with a partner.

  • Widowed and divorced women include all widowed and divorced women irrespective of whether they are living with a partner.

  • Cohabiting women refer to all unmarried women who are living in a cohabiting union.

  • Single women include all unmarried women who are not living in a cohabiting union [11]  Some cohabiting women may have misreported their family...[11].

In Figure 3, we also use information on the year the respondent got married. It is not possible to say if this information relates to a first or higher order marriage. For simplicity, we assume that it relates to first order marriages. In contrast to most other questions in the Mikrozensus, respondents are not required to answer this question. Therefore, the non-response rate for this question is relatively high. In our analysis, we have to omit roughly 10% because of non-response. For the multivariate analysis, we use information on the family form at the date of interview. There is no non-response option on this question.

Educational attainment


There are three binary variables that indicate the highest (post-secondary) diploma obtained at the time of interview:

  • Vocational diploma includes Lehrausbildung, Meister, Fachschulabschluss.

  • College includes a university or a college degree (Universitäts-/Fachhoch-schulabschluss).

  • No diploma includes respondents who did not obtain a post-secondary certificate, i.e. who did not receive a vocational training certificate or a college degree. This category also includes respondents who received nothing more than training on the job (Anlernausbildung). Furthermore, it can include respondents who obtained a primary or secondary school certificate (Hauptschulabschluss, Realschulabschluss, Abitur), provided they did not obtain a vocational training certificate or a college degree.

Activity status


We distinguish:

  • Student;

  • Employed part-time (>0 and <35 working hours);

  • Employed full-time (≥35 working hours);

  • Not working (unemployed and not in the labour force).

The activity status was constructed on the basis of information on the working hours in the reference week (Berichtswoche). This means that we wrongly classify individuals as “not working” who were on sick leave or on holiday during this week. However, using the “current employment status,” which is also recorded, yields artificially high female employment rates. This is because women on parental leave are classified as employed. In the German micro-census, parental leave take-up is unfortunately not surveyed.

Activity status of partner


We distinguish:

  • Student;

  • Employed full-time (≥35 working hours);

  • Not working or employed part-time (>0 and <35 working hours).

East/west German


An east German is a respondent who lived in the five new Länder or East Berlin in 1997. A west German is a respondent who lived in the territories of former West Germany in 1997.

Sample size

Table A –  Sample size mikrozensus 1997

For their valuable comments we would like to thank Johannes Huinink, Heike Trappe and the participants of the conference on “The second demographic transition in Europe”, held in Bad Herrenalb in June 2001. We would also like to thank two anonymous reviewers of Population for their constructive comments and suggestions on an earlier version of this article.


  • Adler M.A., 1997, “Social change and decline in marriage and fertility in Eastern Germany”, Journal of Marriage and the Family, 59(1), pp. 37-49.
  • Andersson G., 1998, “Trends in marriage formation in Sweden 1971-1993”, European Journal of Population, 14(4), pp. 157-178.
  • Beckmann P.,Engelbrech G., 1999, “Beschäftigungsentwicklung und Perspektiven ostdeutscher Frauen in den 90er Jahren”, in E. Wiedemann et al. (eds.), Die arbeitsmarktund beschäftigungspolitische Herausforderung in Ostdeutschland, Nuremberg, Bundesanstalt für Arbeit, pp. 203-226.
  • Bennett N.G., Blanc A.K., Bloom D.E., 1988, “Commitment and the modern union: assessing the link between premarital cohabitation and subsequent marital stability”, American Sociological Review, 53(1), pp. 127-138.
  • Berghahn S., 1992, “Frauenrechte im Vereinigungsprozeß. Das Ende der Privilegien oder sozialer Kahlschlag?”, in C. Faber, T. Meyer (eds.), Unterm neuen Kleid der Freiheit: Das Korsett der Einheit. Auswirkungen der deutschen Vereinigung für Frauen in Ost- und Westdeutschland, Berlin, Ed. Sigma, pp. 63-98.
  • Braun M., Scott J., Alwin D.F., 1994, “Economic necessity or self-actualization? Attitudes toward women’s labour-force participation in East and West Germany”, European Sociological Review, 10(1), pp. 29-47.
  • Brinkmann C., Wiedemann E., 1995, “Arbeitsmarktrisiken im ostdeutschen Transformationsprozess: Ergebnisse des Arbeitsmarkt-Monitors 1989 bis 1994”, Mitteilungen aus der Arbeitsmarkt und Berufsforschung, 28(3), pp. 323-338.
  • Bundesministerium für Arbeit und Sozialordnung (BMA), 2000, “Übersicht über das Sozialrecht”, 6th revised and enlarged edition, Bonn.
  • Council of Europe, 2000, “Recent Demographic Developments in Europe”, Strasbourg.
  • Cromm J., 1998, “Familienbildung in Deutschland: soziodemographische Prozesse, Theorie, Recht und Politik unter besonderer Berücksichtigung der DDR”, Opladen, Westdeutscher Verlag.
  • Deutsches Jugendinstitut, 1993, “Tageseinrichtungen für Kinder. Informationen, Erfahrungen, Analysen”, Munich, Deutsches Jugendinstitut.
  • Deutsches Jugendinstitut, 1998, “Tageseinrichtungen für Kinder. Pluralisierung von Angeboten”, Munich, Deutsches Jugendinstitut.
  • Dorbritz J., 1997, “Der demographische Wandel in Ostdeutschland — Verläufe und Erklärungsansätze”, Zeitschrift für Bevölkerungswissenschaft, 22(2-3), pp. 239-268.
  • Dorbritz J., 2000, “Europäische Fertilitätsmuster”, Zeitschrift für Bevölkerungswissenschaft, 25(2), pp. 235-266.
  • Drobnic S., 2000, “The effects of children on married and lone mother’s employment in the United States and (West) Germany”, European Sociological Review, 16(2), pp. 137-157.
  • Duvander A.Z.E., 1999, “The transition from cohabitation to marriage: a longitudinal study of the propensity to marry in Sweden in the early 1990s”, Journal of Family Issues, 20(5), pp. 698–717.
  • Emmerling D., Riede T., 1997, “Vierzig Jahre Mikrozensus”, Wirtschaft und Statistik (3), pp. 160-174.
  • Engelbrech G., 1997, “Erziehungsurlaub und was dann? Die Situation von Frauen bei ihrer Rückkehr auf den Arbeitsmarkt: ein Ost/West-Vergleich”, IAB Kurzbericht (8).
  • Frerich J., Frey M., 1993, “Handbuch der Geschichte der Sozialpolitik in Deutschland, Band 2: Sozialpolitik in der Deutschen Demokratischen Republik”, Munich, Oldenbourg Verlag.
  • Frerich J., Frey M., 1996, “Handbuch der Geschichte der Sozialpolitik in Deutschland. Band 3: Sozialpolitik in Deutschland bis zur Herstellung der deutschen Einheit”, 2nd edition, Munich, Oldenbourg Verlag.
  • Gysi J., Speigner W., 1983, “Changes in the life patterns of families in the German Democratic Republic”, Berlin (East), Institut für Soziologie und Sozialpolitik an der Akademie der Wissenschaft der Deutschen Demokratischen Republik.
  • Höhn C., 1992, “Population-relevant policies before and after unification of Germany”, in P. Krishnan, C.H. Tuan, K. Mahadevan (eds.), Readings in Population Research, Delhi (India), pp. 1-24.
  • Höhn C., Dorbritz J., 1995, “Zwischen Individualisierung und Institutionalisierung – Familiendemographische Trends im vereinten Deutschland”, in B. Nauck, C. Onnen-Isemann (eds.), Familie im Brennpunkt von Wissenschaft und Forschung, Neuwied, Luchterhand, pp. 149-176.
  • Holst E., Schupp J., 1999, “Erwerbsbeteiligung und Arbeitszeitwünsche 1993 und 1997. Westund Ostdeutschland im Vergleich”, in W. Glatzer, I. Ostner (eds.), Deutschland im Wandel: sozialstrukturelle Analysen, Opladen, Leske und Budrich, pp. 289-306.
  • Huinink J., 1998, “Ledige Elternschaft junger Frauen und Männer in Ost und West”, in R. Metze, et al., (eds.), Der Transformationsprozess: Analysen und Befunde aus dem Leipziger Institut für Soziologie, Leipzig, Leipziger Universitätsverlag, pp. 301-320.
  • Huinink J., 1999, “Die Entscheidung zur Nichtehelichen Lebensgemeinschaft als Lebensform – Ein Vergleich zwischen Ost- und Westdeutschland”, in T. Klein, W. Lauterbach, (eds.),Nichteheliche Lebensgemeinschaften. Analysen zum Wandel partnerschaftlicher Lebensformen, Opladen, Leske und Budrich, pp. 113-138.
  • Huinink J., Wagner M., 1995, “Partnerschaft, Ehe und Familie in der DDR”, in J. Huinink, et al., (eds.), Kollektiv und Eigensinn. Lebensverläufe in der DDR und danach, Berlin, Akademie Verlag, pp. 145-188.
  • Huinink J., Konietzka, D., 2002, “Lebensformen und Familiengründung. Nichteheliche Elternschaft in Ost- und Westdeutschland in den 1990er Jahren”, University of Rostock, mimeo.
  • Kreyenfeld M., Huinink J., 2002, “Der Übergang zum ersten und zweiten Kind. Ein Vergleich zwischen Mikrozensus und Familiensurvey”, Rostock University, mimeo.
  • Kreyenfeld M., 2000, “Changes in the timing of first birth in East Germany after Re-Unification”, Schmollers Jahrbuch – Zeitschrift für Wirtschafts- und Sozialwissenschaften, 120(2), pp. 169-186.
  • Kreyenfeld M., 2001, “Employment and Fertility – East Germany in the 1990s”, University of Rostock, Ph.D. Thesis.
  • Mayer K.U., Diewald M., Solga H., 1999, “Transitions to post-communism in East Germany: Worklife mobility of woman and men between 1989 and 1993”, Acta Sociologica, 42(1), pp. 35-53.
  • Nave-Herz R., 1994, “Familie heute”, Darmstadt, Wissenschaftliche Buchgesellschaft.
  • Obertreis G., 1986, “Familienpolitik in der DDR 1946-1980“, Opladen, Leske und Budrich.
  • Oppenheimer V.K., 1995, “The role of women’s economic independence in marriage formation”, in H.P. Blossfeld (ed.), The New Role of Women, Boulder, Westview Press, pp. 236-243.
  • Peuckert R., 1999, “Familienformen im sozialen Wandel”, 3rd edition, Opladen, Leske und Budrich.
  • Raley K.R., 2001, “Increasing fertility in cohabiting unions. Evidence for the second demographic transition in the United States?”, Demography, 38(1), pp. 59-66.
  • Rindfuss R.R., Brewster K.L., 1996, “Childrearing and fertility”, Population and Development Review, 22 (Supplement), pp. 258-289.
  • Sackmann R., 1999, “Ist ein Ende der Fertilitätskrise in Ostdeutschland absehbar?”, Zeitschrift für Bevölkerungswissenschaft, 24(2), pp. 187-211.
  • Sainsbury D., 1997, “Taxation, family responsibilities, and employment”, in D. Sainsbury (ed.), Gender and Welfare State Regimes, Oxford, Oxford University Press, pp. 185-209.
  • Schimpl-Neimanns B., 1998, “Analysemöglichkeiten des Mikrozensus”, ZUMA-Nachrichten, 22(42), pp. 91-122.
  • Schneider N.F., Matthias-Bleck H. (1999), “Moderne Familie – altes Recht?”, in F.W. Busch, et al., (eds.), Aktuelle Forschungsfelder der Familienwissenschaft, Würzburg, Ergon Verlag, pp. 185-210.
  • Seltzer J., 2000, “Families formed outside of marriage”, Journal of Marriage and the Family, 62(4), pp. 1247-1268.
  • Smock P.J., 2000, “Cohabiting in the United States. An appraisal of research themes, findings, and implications”, Annual Review of Sociology, 26, pp.
  • Statistisches Bundesamt, 2001a, “Bevölkerung und Erwerbstätigkeit. Gebiet und Bevölkerung 1999.” Fachserie 1, Reihe 1, Stuttgart: Metzler-Poeschel.
  • Statistisches Bundesamt, 2001b, “Altersspezifische Geburtenziffern für Ost- und Westdeutschland für das Jahr 2000”, personal communication from Dieter Emmerling (Statistisches Bundesamt).
  • Statistisches Bundesamt, 2001c, “Sozialleistungen. Tageseinrichtungen für Kinder 1998”, Stuttgart.
  • Trappe H., 1995, “Emanzipation oder Zwang? Frauen in der DDR zwischen Beruf, Familie und Sozialpolitik”, Berlin, Akademie Verlag.
  • Wirth H., 2000, “Bildung, Klassenlage und Partnerwahl”, Opladen, Leske und Budrich.
  • Witte J.C., Wagner G.G., 1995, “Declining fertility in East Germany after unification: a demographic response to socioeconomic change”, Population and Development Review, 21(2), pp. 387-397.


[*] Institute for Sociology and Demography, University of Rostock.

[**] Max Planck Institute for Demographic Research, Rostock.

[1] From 1961, women were allowed to take one year’s unpaid leave after childbirth. In 1976, mothers with two and more children became eligible for one year’s paid leave; the maternity allowance was equal to the usual sick pay granted after the 7th week of sickness. From 1984, mothers with more than two children were eligible for 18 months of paid leave. In 1986, paid leave was extended to all mothers. If no day care place could be found, all mothers were entitled to extend their period of unpaid leave up to the child’s third birthday (Frerich and Frey, 1993; Cromm, 1998).

[2] At German unification, the Unification Treaty (Einigungsvertrag) that came into force prescribed that the institutions of East Germany would be replaced by West Germany’s. It should be noted, however, that some of the former East Germany’s provisions were abolished only gradually. Paid leave when a child was sick remained in force until July 1991. The rules on parental leave and child benefits were changed in January 1991 (Berghahn, 1992, pp. 78 seq.; Frerich and Frey, 1996).

[3] Since 1992, married women (and in principle also married men) are entitled to take 10 days of paid leave to care for a sick child (20 days for parents with more than one child). Single parents may take 25 days of paid leave to care for a sick child (50 days for single parents with more than one child) (BMA, 2000, pp. 152 seq.).

[4] One of the main reasons for public day care coverage remaining high is probably the reduction in demand for care eventually brought about by the low birth rates after unification. In other words, although day care slots were cut after unification, because the number of children declined even faster, the availability ratio could be kept at a high level (for a fuller discussion of this issue, see Kreyenfeld, 2001).

[5] The analysis was actually conducted with data from the scientific use file of the micro-census, which is a 70% sample of the original micro-census.

[6] According to calculations based on the Familiensurvey 2000, less than 5% of the children of women aged 31-35 have left the parental home. Less than 1% of their children were adopted or stepchildren (Kreyenfeld and Huinink, 2002).

[7] For this analysis, we use retrospective information on the date of marriage (see Appendix for details).

[8] Otherwise, it would have been more appropriate to employ event history techniques on the transition to first marriage (see e.g. Andersson, 1998).

[9] We also omitted respondents for whom information on educational attainment was missing (for details, see Table A in the Appendix).

[10] This implies that couples anticipate the woman’s future employment behaviour when deciding to get married. We are, however, unable to unravel the temporal sequence of the withdrawal from the labour force, the birth of the first child and the marriage decision. Even if we had this information, however, it is doubtful that it would provide a deeper insight into the causal mechanisms behind the decision to get married. Although this decision might be motivated by the intention to become a housewife and mother, a large number of women are nonetheless childless and in full-time employment at the time of getting married.

[11] Some cohabiting women may have misreported their family status. As discussed in Part II, transfer payments are means tested and the cohabiting partner’s income is included in the assessment. Cohabiting women who are receiving social benefits may report themselves as being single, out of fear that declaring the cohabiting partner would have negative consequences for benefit eligibility. However, we have no information on this issue.



In contrast to West Germany, where marriage and childbirth have been strongly coupled, the German Democratic Republic (GDR) displayed high rates of non-marital childbearing. Researchers attributed this pattern to “misguided” GDR family policies that encouraged women to remain unmarried after childbirth. With German unification, East Germany’s legal and political institutions — including family policies — were replaced by those of West Germany. Against this background, it was widely expected that east German non-marital birth rates would soon fall to west German levels. After unification, however, they increased even further.
This article argues that the enormous east-west differences in non-marital childbearing in the 1990s can be attributed to differences in women’s work orientation. Despite unfavourable labour market constraints and social policies that encourage women’s withdrawal from the labour force after childbirth, east German women, compared with their west German counterparts, are still more likely to be in full-time employment, and to re-enter the labour force sooner after childbirth. Our empirical investigation, drawing on data from the German 1997 micro-census, reveals a strong effect of women’s education and employment on marriage in west Germany, whereas in east Germany the probability of living in a marital union is hardly correlated at all with women’s employment characteristics. We conclude that a generally strong female work orientation and the wide availability of public day care facilities are the most important factors weakening the economic incentives for east German women to get married at childbirth.


  1. Family policies and non-marital birthsin east and west Germany
    1. Family policies before unification
    2. Family policies after unification
    3. Female employment in east and west Germany
  2. Data source and method of analysis
    1. Data source
    2. Method of analysis
  3. Empirical analysis
    1. “Unmarried parent”: a stable family status?
    2. Marital union versus cohabiting union
    3. East-west comparison
  4. Conclusion

To cite this article

Dirk Konietzka, Michaela Kreyenfeld, “  ”, Population 2/2002 (Vol. 57) , p. 331-357
URL : www.cairn.info/revue-population-2002-2-page-331.htm.
DOI : 10.3917/popu.202.0359.

Previous Pages 331 - 357 Next
© 2010-2014 Cairn.info
back to top

Message sent

Your email address has been saved.
We will notify you when this article becomes available in English.