Determinants of and Obstacles to Dual Careers in Germany

Full text

Page 1

Zeitschrift für Familienforschung, 19. Jahrg., Heft 3
Alessandra Rusconi & Heike Solga
Determinants of and obstacles to
dual careers in Germany
Determinanten und Hindernisse für Doppelkarrieren in Deutschland
Abstract
Education expansion has led to an increasing
proportion of couples in which both partners
hold an academic degree. Although in these
couples the potential for dual careers is high,
their realization remains a great challenge,
mostly due to restrictions on female careers.
We argue that the chances of achieving dual
careers depend not only on individual char-
acteristics of each partner, but also on the
constellation within couples with regard to
age, profession, and child-rearing. We there-
fore look first at different characteristics
within German academic couples and ana-
lyze how these enhance or constrain the
chances of achieving dual careers. Then, we
examine recruitment practices at German
universities, especially their sensitivity re-
garding dual-career issues. Our results show
that child rearing contributes to the risk of
couples having a one-career coordination
strategy and that gender-blind recruitment
criteria contribute to a large extent to this
risk. We further test two common explana-
tions of the hindrance of dual careers and
show that neither are women generally dis-
advantaged, nor does the age difference in
couples gender-neutrally define the older
partner’s career as the leading one.
Zusammenfassung:
Die Bildungsexpansion hat dazu beigetragen,
dass bei einem zunehmenden Anteil der Paa-
re beide Partner über einen akademischen
Abschluss verfügen. Obgleich diese Partner-
schaften ein hohes Potential für Doppelkar-
rieren besitzen, so bleibt dessen Realisierung
eine große Herausforderung – oft aufgrund
von Einschränkungen weiblicher Karrieren.
Wir zeigen, dass nicht nur individuelle
Merkmale der beiden Partner die Realisie-
rungschancen von Doppelkarrieren beein-
flussen, sondern auch Konstellationen inner-
halb der Partnerschaft im Bezug auf Alter,
Beruf und Kinder. Mit Blick auf Akademi-
kerpaare untersuchen wir zunächst, wie un-
terschiedliche Paarkonstellationen die Reali-
sierungschancen von Doppelkarrieren för-
dern oder behindern. In einem weiteren
Schritt untersuchen wir Einstellungsprakti-
ken an deutschen Hochschulen und deren
Sensibilität gegenüber Doppelkarrieren. Un-
sere Ergebnisse zeigen, dass Kinder das Ri-
siko bei Paaren erhöhen, nur eine Karriere zu
verfolgen, und dass geschlechtsblinde Ein-
stellungskriterien zu diesem Risiko beitra-
gen. Zudem werden wir die zwei herkömm-
lichen Erklärungen zur Behinderung von
Doppelkarrieren überprüfen und dabei zei-
gen, dass weder Frauen generell benachtei-
ligt sind, noch dass ein Altersunterschied der
Partner den Karrierevorrang des älteren
Partners geschlechtsneutral definiert.

Page 2

312Rusconi & Solga: Determinants of and obstacles to dual careers in Germany
Key words: dual careers, academic couples,
Germany, family relationships, university,
employment, careers
Schlagworte: Doppelkarrieren, Akademiker-
paare, Deutschland, familiäre Beziehungen,
Universität, Hochschulen, Beruf, Karriere
1. Introduction
For a long time, the gender gap in career chances connected with a hierarchical one-
career pattern in couples was explained by differences in educational degrees at-
tained by men and women. Yet, in many Western countries educational expansion
has resulted in an enormous reduction of educational gender differences. In 2004
about 19.5 percent of German men and 14.5 percent of women – aged 30 to 49 –
held an academic degree.1 And not only did the number of highly qualified indi-
viduals – and especially women – increase; there was also considerable growth in
the number of couples in which both partners hold an academic degree (in the fol-
lowing: “academic couples”). In Germany today, one third of the men holding a
university degree live with an academically-trained woman, and almost half of the
female academics live together with a male academic (cf. Table 1). Dual-career
couples – couples in which both partners pursue a professional career – are none-
theless quite seldom in Germany.
In academic couples, both partners have made substantial training investments
and often show a pronounced interest in professional careers – leading to a high
potential for entwining individual professional careers into dual careers. Yet dual
careers still remain a great challenge: often dual-career efforts fail, most frequently
because of restrictions on female professional careers. Among the academic women
aged between 30 to 49 years, about 38 percent were unable to translate their aca-
demic degree into a (full-time or part-time) professional job in 2004. These women
were either not employed or (only) working in non-professional jobs. Among the
academic men only 22 percent could not do so. That the interest in professional ca-
reers by male and female academics is increasing, however, is indicated by the large
and, for men, rising proportion of academics not living with a partner. In the career-
intensive period between ages 30 to 50, every third academically trained man and
woman is not “committed” to living together with a partner (see Table 1). Between
1971 to 2004, for men this figure increased from only 11 percent to 27 percent, in-
dicating the decreasing chances of men to find “traditional” women for a partnership
and/or their increasing interest to first establish their career before entering a (dual-
career) family.

1 In 1971, 7 percent of the men and less than 2 percent of the West German women held an
academic degree.

Page 3

Zeitschrift für Familienforschung, 19. Jahrh., Heft 3/2007, S. 311-336 313
Table 1: Qualification levels of partners of academically trained men and women
in 1971, 1997, 2004 (in rounded percentages)
Partner holding an academic
degree (“academic couples”)
Partner without an
academic degree
No partner*
1971**
72
13
Men
Women
15
50
11
33
1997
47
20
Men
Women
27
46
25
34
2004
39
23
Men
Women
Case selection: German academics aged between 30 and 49 years.
Difference to 100 percent = missing information on partner’s qualification.
Due to data restriction:
* this category includes women and men – aged between 30 and 49 years – without partners
and in living-apart-together-arrangements.
** this year only includes married couples.
Source: Own calculations, German Micro-census 1971, 1997, 2004, weighted by individual
projection factor.
33
47
27
30
Although the number of academic couples increased from 1.1 percent of all couples
in 1971 to about 9 percent in 2004, the one-career-pattern is still common in Ger-
many. Even in academic couples, often only the male partner pursues a professional
career. In 2004 in about every third academic couple, the woman was either not em-
ployed or did not pursue a professional job, despite her academic degree (cf. section
3).2
The goal of this article is to reveal alternative explanations for this gendered ca-
reer gap that go beyond the educational-career resource argument. Therefore, we
study career chances of academic couples, i.e., we keep the (initial) educational re-
sources between the two partners equal. This allows us to explore further determi-
nants and obstacles for dual careers of German academic couples. In the following,
we theoretically outline and empirically show that dual-career chances depend not
only on individual characteristics of each partner, but also on the constellation
within couples with regard to age, profession, and child-rearing (sections 2 and 3).
We then continue with a case study of employers’ perspective on dual careers and a
description of obstacles and constraints to dual careers caused by recruitment prac-
tices of German higher education institutions (section 4). In the concluding section,
we summarize the main constraints and determinants of dual careers among aca-
demic couples.

2 Own calculations based on the Micro-census 2004 (cf. information at Table 1).

Page 4

314Rusconi & Solga: Determinants of and obstacles to dual careers in Germany
2. Determinants of dual careers – Theoretical
considerations on the presence of children, couples’ age
difference and professional constellation
A large body of literature – especially on gender differences in career opportunities
– focuses on how individual features (such as age, gender, educational level, profes-
sional field and experience) influence employment and career chances (e.g. Geenen
1993; Lauterbach 1994; Stroh & Reilly 1999; Allmendinger, Fuchs & Stebut 2000;
Blossfeld & Drobnič 2001; Born & Krüger 2001; Baecker 2003). However, scholars
have paid far less attention to the couple as unit of analysis and thus underestimated
the importance of the constellation of the partners’ individual characteristics within
the couple. In this article, we therefore favor such a relational approach in which
not only the individual age and professional field of the two partners as career de-
terminants are taken into account, but the age and professional constellation of the
two partners as well. In addition, we investigate the effect of children in order to re-
veal whether childlessness is a “dual-career strategy” to reduce family duties for
both partners. Before presenting our empirical findings (in section 3), we will out-
line how the age difference between partners and couples’ professional heterogene-
ity or homogeneity may influence dual-career opportunities.
In the literature, the age difference between partners is often seen as a gender-
neutral determinant of the career chances of the two partners. This explanation appears
to be straightforward on first glance: careers and career steps take time, and if the age
of the two partners is unequal, unequal time periods have been available for each of
their respective careers. Given the age difference between the partners and the result-
ing differences in career time, the older partner typically has already proceeded further
in his/her career. According to the New Household Economics, it is therefore in the
household’s common interest for the older partner to have the “primary” career during
a significant proportion of the life course (e.g. Hawkes, Nicola & Fish 1980; Kalter
1998; Becker & Moen 1999). A similar argument is made by resource-exchange theo-
ries, here the older partner has already accumulated more resources to assert his/her
interests (e.g., Blood & Wolfe 1960). Since typically women are younger than their
male partners (see Table 2), men will have the “primary” and women the “secondary”
career.3 Correspondingly, one of the dominant explanations in the literature for the
hindrance of female career efforts and thereby of dual careers addresses the age con-
stellation, but from a gender-neutral perspective.
Here, we will add a gendered perspective on how the age constellation affects
dual-career opportunities of academic couples. Age is not only a natural, but also a
socially constructed characteristic. According to the naturalization of social relation-
ships (cf. Douglas 1991), couples’ age constellation may also define hierarchical
and power relations between partners – i.e., the “older” partner is seen as the more
experienced individual, legitimating his (or, less often, her) more powerful position
in the couple. The typical age difference between men and women in couples may

3 The mean age difference between partners is about 3 years, which appears being a quite
stable feature over time and across countries.

Page 5

Zeitschrift für Familienforschung, 19. Jahrh., Heft 3/2007, S. 311-336315
therefore “naturalize” hierarchical gender roles. In other words, different types of
gender relations might be hidden behind or connected to the age difference between
partners. The choice of the “typical” age constellation in heterosexual couples – i.e.,
the choice of an older man or a younger woman as a partner – may indicate a prefer-
ence for hierarchical gender relations rather than simply a temporal careeer advan-
tage of the man; whereas the choice of an “atypical” age constellation – i.e., the
choice of a younger man or an older woman as a partner – might express a more
egalitarian gender relationship between the two partners.
Table 2: Age constellations in German academic couples, 2004
(in rounded column percentages)
TotalFemale partner is
30–39 yrs old
Female partner is
40–49 yrs old
a) Typical age constellations
– No age difference: Age difference is max. 2 years
(man is up to 2 yrs older/younger than the woman)
–Man is 3 to 7 yrs older than the woman
–Man is 8 or more yrs older than the woman*
52 53 52
32
10
34
9
31
10
b) Atypical age constellation
–Woman is 3 or more yrs older than the man
Total
Case selection: Heterosexual couples, only German citizens, women between 30 and 49 years
old, men born after 1943.
* The mean age difference in these couples is 11 years.
Source: Own calculations, German Micro-census 2004, weight by family projection factor.
6
100
4
100
7
100
In both (relational) perspectives, it is not only the individual age that may affect
one’s own career opportunities, but also his/her age relative to his/her partner’s age.
A gender-neutral effect of the age constellation would be indicated if the older part-
ner – irrespective of gender – has a head start in terms of career and enjoys the
“leading” career in comparison to the younger partner. Thus, dual careers should be
more frequently found in couples in which the two partners are of similar age than
in couples with a pronounced age difference. In contrast, according to a gender per-
spective on age constellation we would expect that couples in which the woman is
older than the male partner more often have egalitarian arrangements. Thus, they
should more frequently realize dual careers than age-typical couples.
Past research has also shown that professional fields are an important determinant
of career opportunities. Their influence is connected to the relationship between labor
market supply and demand at a given time as well as field-specific career-pathway
patterns. Additionally, whether the professional fields of the two partners define occu-
pational homo- or heterogeneity is of special importance. Firstly, in small local labor
markets, partners in occupationally homogeneous couples might have to search for a
job with the same employer. If (un)official anti-nepotism regulations exist, occupa-
tional homogeneity might constrain the professional career of (at least) one of the two
partners and thereby hinder dual careers. Secondly, applying Parsons’ idea (1959) of
stability of and competition within couples, occupational homogeneity might lead to
greater competition between partners, for example with regard to mobility decisions

Page 6

316Rusconi & Solga: Determinants of and obstacles to dual careers in Germany
and the division of household (family) duties, and therefore restrict the realization of
dual careers for the sake of the “relationship”. Alternatively, occupational homogene-
ity understood as shared social capital might instead constitute an opportunity for ad-
vantageous cooperation or mutual understanding for career steps and demands and
thereby support dual careers (for a discussion see Bernasco, de Graaf & Ultee 1998;
Sonnert 2005). Competition between partners or limitations in labor markets would be
indicated if partners working in the same occupational field have a lower probability of
a dual career than heterogeneous couples, whereas increased cooperative relationships
through occupational homogeneity would be corroborated if dual careers exist more
frequently in homogenous couples.
Finally, the presence of children as well as their age(s) are known to influence
females’ professional careers and, therefore, should determine dual careers as well.
Dual-career couples with children not only face the challenge of how to combine
two careers, but also of how to intertwine occupational demands with childcare du-
ties (e.g. Adler et al. 1989; Moen & Wethington 1992; Nock 1998). Several scholars
note that many couples often start on an egalitarian basis, but after the birth of a
child – or after mobility decisions for the benefit of the male career – they move to-
wards a more traditional division of labor within the family (Gilbert 1985; Hensel
1991; Levy & Ernst 2002). This shift reflects still prevalent gender roles that also
apply to academic women according to which, after childbirth, female partners
should take on the main burden of child-rearing and show greater flexibility in order
to accommodate family duties, whereas male partners are expected to ensure the fi-
nancial resources of the family. As a consequence, the chances for female partners
to purse their own careers diminish, since even temporary career compromises often
have enduring negative consequences for professional careers.4
This impact of children may vary, however, with employment-family-coordination
patterns of couples. Past research has identified three main patterns. The first one is the
hierarchical (traditional) model in which only one partner – most often the man – has
the dominant career role, while the other partner supports (t)his leading career through
primary responsibility for “private/family matters”. Explanations for the wide preva-
lence of this hierarchical relationship between partners are gender role ideologies
(Bielby & Bielby 1992), considerations by the literature on new household economics
(Becker 1991) and resource-exchange explanations (e.g. Blood & Wolfe 1960; Emer-
son 1976). In the individualistic model, each of the two partners independently pursues
his/her own career and the partnership itself plays a secondary role. This arrangement
is frequently linked with long-distance or commuter living arrangements, with the goal
of optimally structuring the career chances of both partners (Kilpatrick 1982). How-
ever, as soon as children are born, this individualistic arrangement begins to waver, as
it is based above all on the absence of the male’s resistance with regard to the female
career – and this only as long as his own career is not jeopardized by it (Hertz 1986;

4 Research has shown, however, that due to different professional cultures and the avail-
ability of alternative career patterns, professions vary in their potential for the achievement
of a work/life balance and, thus, the realization of dual careers (for medicine and psychol-
ogy in Germany, see Hoff et al. 2002; Dettmer & Hoff 2005).

Page 7

Zeitschrift für Familienforschung, 19. Jahrh., Heft 3/2007, S. 311-336317
Levy & Ernst 2002).5 Public and private childcare and external household assistance
are mostly considered as paid substitutions for the woman. If the costs are high, the
risk of a re-traditionalization, i.e., a transition to the hierarchical pattern increases.
Common strategies by academic women to avoid such a collision between family and
career cycles, are to remain childless (Gilbert 1985) or to postpone having children
(Austin & Pilat 1990; Costa & Kahn 2000; Gappa 1980; Monk-Turner & Turner
1987).6 The third coordination pattern, the egalitarian model, is considerably less fre-
quent than the other two patterns. Here, professional careers and the family are of
equal importance to the partners. Both partners therefore make compromises in their
careers (or are willing to do so) in favor of their “relationship” and career opportunities
for both partners (e.g. Becker & Moen 1999; Hardill et al. 1999; Costa & Kahn 2000).
With the data at hand, we are unable to observe directly couples’ coordination
strategies and, thus, their influence on dual careers. Using our idea of a gendered age
constellation in couples (see above), however, we are able to indirectly derive indica-
tions on couples’ career-household strategies. A gendered influence of the age con-
stellation would be indicated by a “positive” main effect of the age-atypical constella-
tion, i.e. age-atypical couples more often realize dual careers than age-typical couples.
Given that there is a significant main effect of couples’ age constellation, the results of
models that simultaneously control for children could signal the following: a) if the
main effect disappears after controlling for children then age-atypical couples are more
often childless – indicating (indirectly) that they follow more often the individualistic
model than do age-typical couples; b) if the main effect of couples’ age constellation
remains unchanged after controlling for children, then age-atypical couples realize
dual careers to a larger extent than age-typical couples, and that regardless of children.
This might indicate that age-atypical couples more often follow an egalitarian coordi-
nation pattern. Furthermore, we will control for differences between birth cohorts, in
order to determine whether the relevance of individualistic and/or egalitarian coordi-
nation patterns has risen in recent years.
3.Empirical findings
In this section we empirically analyze the impact of the age constellation, occupational
homogeneity and children on dual careers. We use the German Micro-census of 2004
(see methodological appendix).7 Our focus is on heterosexual academic couples, de-

5 Although highly qualified men more frequently show an egalitarian attitude to the work of
women and division of labor in the household, in practice they rarely show egalitarian be-
havior (Hardill et al. 1999).
6 According to a study on academic careers (Krimmer et al. 2003), 26 percent of present-
day German female professors consciously elected not to have children; this figure was
only 10 per cent for male professors. As a result, in Germany today 50 per cent of female
professors, but only 20 per cent of male professors are childless.
7 With this cross-sectional survey dual careers cannot be reconstructed in terms of decision
situations, negotiation processes and decision outcomes. However, there is currently no
alternative data base for investigating academic couples. In existing longitudinal studies,
such as the Socio-economic panel of the German Institute for Economic Research and

Page 8

318Rusconi & Solga: Determinants of and obstacles to dual careers in Germany
fined as couples in which both partners hold an academic degree. Due to limited career
information, we have defined dual-career couples as couples in which both partners
carried out a professional occupation at the time of the survey (2004).8 We have de-
fined professional jobs as those that formally require an academic degree or are usually
occupied by persons holding an advanced academic degree.
According to this (broad) definition, half of the academic couples had dual-career
arrangements, however, a significant proportion of these dual-career couples fol-
lowed the “traditional” working-time pattern, i.e., the woman worked part-time (see
Table 3).9 Although both partners held an academic degree, 27 percent of the cou-
ples have a (traditional) hierarchical career arrangement: the man was pursuing a
professional occupation, while the woman was either not employed (14 percent) or
not carrying out a professional occupation (13 percent).
Table 3:Employment/Career constellation depending on age of the woman and
presence of underage children* in academic couples, 2004 (in rounded
row percentages)
Dual Careers No Dual Career
Both full-
time
Man full-
time/Woman
part-time
Unusual
DC
Man ca-
reer/Wo-
man job
Man career/
Woman not
employed
Woman
career/
Man job
Woman ca-
reer/Man
not em-
ployed
Other
30-39 yrs (Total)
No children
With children
40-49 yrs (Total)
No children
With children
Total
Case selection: Heterosexual couples, only German citizens, women between 30 and 49 years
old, men born after 1943.
* The Micro-census’ category “children” includes biological, adopted, foster and stepchil-
dren insofar as they live together with at least one parent in the household.
Source: Own calculations, German Micro-census 2004, weight by family projection factor.
31
52
19
24
39
18
27
18
6
24
23
10
29
21
3
2
3
2
2
3
3
11
10
12
14
16
13
13
16
6
21
12
5
16
14
7
11
6
9
11
9
9
2
3
2
2
4
2
2
11
9
12
12
13
11
11
Our analyses will take place in three steps. As an essential precondition of dual ca-
reers we investigate first whether both partners are employed – regardless of the
qualification level of jobs. As a second precondition we examine whether the female

German Life History Studies of the Max Planck Institute for Human Development, the
number of academic couples is too small.
8 The classification of professional jobs is based on the classification of occupations by the
German Federal Statistical Office (it is a 3-digit code comparable to the International
Standard Classification of Occupations, ISCO).
9 In the last 10 years, the chances of realizing dual careers have not changed. In 1997, only
half of the academic couples had a dual-career arrangement; in 29 percent of the couples
both partners pursued a professional job full-time and another 19 percent of the couples
followed a traditional working-time arrangement (man full-time – woman part-time) (cf.
Solga, Rusconi & Krüger 2005).

Page 9

Zeitschrift für Familienforschung, 19. Jahrh., Heft 3/2007, S. 311-336319
partners have access to professional jobs. And finally – as a “result” – we estimate a
model on dual careers. For all three steps we present results of multinomial logistic
regressions, as each of the dependent variables has several categories. In the regres-
sion tables, we display the odds ratios. Odds ratios greater than 1 signify a greater
chance/risk, odds ratios smaller than 1 indicate a lower chance/risk in comparison to
the respective reference category (which has the value 1).
Starting with the first precondition for dual careers, the employment of both part-
ners, we note that the overwhelming majority of the academic couples realize dual-
earner arrangements (79 percent). Only 17 percent of the couples follow the male-
breadwinner model or the hierarchical coordination pattern (i.e., only the male partner
is employed). However, Table 4 shows that female employment – particularly full-time
– is strongly related to the presence of underage children in the family household.
Among childless academic couples only a minority follows a traditional employment
constellation in which the woman is either not employed (6 percent) or has only a part-
time job (15 percent). In over two thirds of these childless couples both partners are
employed full-time. Quite to the contrary, less than one third of couples with underage
children shows a full-time dual-earner arrangement. A far more common strategy for
academic couples with children is the part-time employment of the female partner (40
percent) or her complete withdrawal from the labor market (22 percent). Thus, once
children are born, even couples in which both partners have made considerable educa-
tional investments are characterized by a more traditional (gendered) division of labor.
By reducing the female commitment to the labor market, this traditional arrangement
may also hinder dual careers because professional careers are seldom achieved in part-
time positions and/or with (temporal) interruptions in employment careers.
Table 4: Employment constellation depending on age constellation and presence
of underage children in academic couples, 2004 (in rounded row
percentages)
Both
full-time
Man full-time/
woman part-
time
Only man
employed
Unusual
employment
constellation*
No children
Total
No age difference (max. 2 yrs)
Man is 3-7 yrs older
Man is 8 or more yrs older
Woman is 3 or more yrs older (= atypical)
With children
Total
No age difference (max. 2 yrs)
Man is 3-7 yrs older
Man is 8 or more yrs older
Woman is 3 or more yrs older
(= atypical)
* Both partners work part-time; woman full-time/man part-time; only woman employed; both
partners not employed.
Case selection: Heterosexual couples, only German citizens, women between 30 and 49 years
old, men born after 1943.
Source: Own calculations, German Micro-census 2004, weighted by family projection factor.
69
70
67
68
72
15
15
15
15
9
6
6
7
6
10
10
9
11
11
10
30
29
31
33
31
40
41
40
40
39
22
22
22
19
22
7
8
7
8
8

Page 10

320Rusconi & Solga: Determinants of and obstacles to dual careers in Germany
Noteworthy, 29 percent of the dual-earner academic couples show an occupational
homogeneity; i.e., both partners are employed in the same occupational sector. With
regard to the age constellation, atypical couples (i.e., in couples in which women are
the older partner) do not differ in their employment arrangement from couples with
a typical age constellation.
By means of multivariate analyses we will now determine the (net) influence of
the age constellation, occupational homogeneity and children on the chances of re-
alizing dual-earner arrangements. The findings presented in Table 5 (model 1 and
2) disclose that for the classical gendered division of labor not children as such but
underage children prove to be of tremendous importance, particularly if they are
younger than three years old. In comparison to childless couples, in families with
small children the female partners are twice more likely to be not employed than to
purse at least a part-time job, while their male partners are full-time employed.
Thus, even in academic couples, children hinder an egalitarian employment con-
stellation. And this negative effect of children on dual-earner arrangements does not
vary among couples with different age constellations (data not shown).10
The age constellation of couples also has an influence on the employment ar-
rangements. Yet, as revealed by the interaction terms, this effect varies by women’s
birth cohort (cf. model 2 in Table 5). For the younger cohort (i.e., women aged be-
tween 30 and 39), in couples in which the partners are of the same age, the question
appears to be either all or nothing: either both partners have a full-time job (odds
ratio = 1,68), or the woman forgoes formal employment altogether (odds ratio =
1,58).11 Thus, female part-time employment does not appear a “favored” strategy for
these couples. In terms of full-time or part-time dual-earner arrangement, both age
atypical couples (woman is older) and those with a very typical age constellation
(i.e., woman is 3 to 7 years younger) are not different from same-aged couples. In
contrast, couples in which the man is 8 or more years older than the woman have a
lower chance of following a full-time dual-earner arrangement than couples of other
age constellations.12 In terms of traditional employment arrangements (“only the
man is employed”), however, couples in which male partners are 8 years older than
the women are not different from same-aged or age atypical couples. In addition,
couples with a very typical age constellation show even lower odds for this tradi-
tional division of labor than couples of other age constellations, but also “unusual”
employment constellations are found less often among such traditional couples.13

10 For space reasons the results of the model with interaction effects between age constella-
tion and children is not included in Table 5. Data is available from the corresponding
author on request.
11 Because of the included interaction effects (in model 2) the odds of employment pattern of
same-aged couples (reference category) are displayed by the main effects of “age of the
women” (or women’s birth cohort).
12 Odds ratio difference = 1.43*0,49 = 0,7
13 For “only man employed” the odds ratio difference for couples in which the man is 3 to 7
year older than the woman is 1,23 (n.s.) *0,70 = 0,86 (n.s. means not significant); for “un-
usual employment constellation” the odds ratio difference is 1,37*0,54 = 0,74.

Page 11

Zeitschrift für Familienforschung, 19. Jahrh., Heft 3/2007, S. 311-336321
Table 5: Multinomial logistic regressions on the employment constellation in
academic couples (reference category: Man full-time – Woman
part-time) (odds ratios, N = 4,308 couples)
Both full-time
employed
Only man em-
ployed
Unusual constel-
lation+
Mod. 1Mod. 2
1
1,06
1,43**
1,21
1
1,68***
1
0,30***
0,18***
0,14***
1
1,08
0,60***
Mod. 1
1
1,05
1,04
1,33
1
1,44***
1
0,71*
0,66**
1,37
1
2,15***
1,46***
Mod. 2
1
1,23
0,98
1,18
1
1,58***
1
0,71*
0,66**
1,36
1
2,17***
1,49***
Mod. 1
1
1,05
1,10
1,07
1
0,97
1
0,36***
0,25***
0,25***
1
1,77**
1,07
Mod. 2
1
1,37*
1,31
0,88
1
1,20
1
0,36***
0,25***
0,25***
1
1,80**
1,10
Age constellation (Ref: No age difference, max. 2 yrs.)1
Man is 3-7 yrs older
Man is 8 or more yrs older
Woman is 3 or more yrs older
Age of the woman (Ref: 40-49 yrs old)
30-39 yrs old
Number of children (Ref: No children under 18 yrs)
1 child
2 children
3 or more children
Age of the youngest child (Ref: no child under 11 yrs) 1
Under 3 yrs
3-10 yrs
Interaction: Age of woman * Age constellation
30-39 yrs old * Man is 3-7 yrs older
30-39 yrs old * Man is 8 or more yrs older
30-39 yrs old * Woman is 3 or more yrs older
1,00
1,07
1,20
1
1,49***
1
0,30***
0,18***
0,14***
1,09
0,60***
0,85
0,49**
1,18
0,70*
1,03
1,62
0,54**
0,64
1,95
Improvement of fit (df): 1249,84 (42) *** (Model 1)
1267,65 (51) *** (Model 2)
+ Unusual employment constellation: Both work part-time; woman full-time/man part-time;
only woman employed; both partners not employed.
Controlled for size of residence, place of residence in East or West Germany, mar-
ried/unmarried cohabitation
Coefficients are significant: p<0.01***; p<0.05**, p<0.1
Case selection: Heterosexual couples, both partners German citizens, women between 30 and
49 years old, men born after 1943.
Source: Own calculations, German Micro-census 2004, unweighted.
For the older cohort (i.e., women aged between 40 and 49), differences between
couples are less pronounced. In contrast to the younger cohort, “unusual” employ-
ment constellations are found surprisingly to a higher extent in age-typical couples
(odds ratio = 1,37) than in couples with other age constellations. In addition, if their
partner is 8 or more years older, the 40-to-49-year-old female partners have a
greater probability to work full-time than to follow a part-time dual-earner arrange-
ment (odds ratio = 1,43). In sum, in both cohorts, contrary to our expectations, age-
atypical couples do not have a higher probability of dual-earner arrangements.
All together, underage children lower the chances of female academics’ labor
force participation and thereof dual-earner arrangements. Moreover, our results in-
dicate that this employment constellation varies by both women’s birth cohort and
couples’ age constellation. In academic couples belonging to the younger cohort,
men are more likely to enjoy an employment advantage. More or less independently
of couples’ age constellation, in this cohort there exists an employment “divide” be-
cause full-time dual-earner arrangements or the male-breadwinner model are more
common than the part-time dual-earner model. In the older cohort, an older male

Page 12

322Rusconi & Solga: Determinants of and obstacles to dual careers in Germany
partner appears to enhance the chances of a dual full-time arrangement for couples
in which women are 8 years younger than the men and of unusual employment con-
stellations for age-typical couples (man 3 to 7 years older). These variations in the
impact of the age constellation by women’s birth cohorts are not caused by (a com-
position effect with regard to) children, as these were simultaneously taken into ac-
count in the models (see also Appendix, Table A3). Therefore another reason may
lie behind this cohort-specific influence of the age constellation. The variation might
indicate that when the woman is still young and in the process of positioning herself
on the labor market, more couples might give priority to the employment of the
(older) already more established male partner. Once couples and thus women age,
they might (be able to) develop more egalitarian employment arrangements. In con-
trast to this age-effect explanation, there might be also a cohort explanation accord-
ing to which women and men of the 1955-to-1964-cohort, for example, have more
egalitarian gender roles or had better career (labor market) conditions at the begin-
ning of their professional careers (i.e., in the 1980s). Given the limitations of the
cross-sectional data used here, we are unable to differentiate between the two expla-
nations.
As labor market research has shown, a second major constraint for dual careers
are the career chances of women. Table 6 presents the results for the influence of the
age constellation, occupational homogeneity and children on the chances of em-
ployed female partners (in academic couples) to follow a professional occupation.
Similar to the employment constellation, the presence of underage children reduces
the probability of female full-time professional employment (indicated in model 1
by the odds ratios for women with children for the category “full-time professional
job” which are smaller than 1). This negative effect of children on female full-time
professional engagement does not vary between couples with different age constel-
lations (data not shown).14 Academic women with children do however, have a
greater likelihood in comparison to childless women, of being able to at least pursue
a professional job part-time (indicated in model 1 by the odds ratios for women with
children for the category “no professional job” which are smaller than 1). Thus,
those women who succeed in being employed “in spite of” children are then more
likely to pursue a professional occupation (at least part-time) than just a “job”.

14 For space reasons the results of the model with an interaction effects between age con-
stellation and children is not included in Table 6. Data is available from the corresponding
author on request.

Page 13

Zeitschrift für Familienforschung, 19. Jahrh., Heft 3/2007, S. 311-336323
Table 6: Multinomial logistic regressions on the chance of being employed in a
professional job by female academics in academic couples (reference
category: Part-time professional job) (odds ratios, N = 3,538 only
employed female academics)
No professional jobFull-time
professional job
Mod. 1Mod. 1 Mod. 2Mod. 2
Age constellation (Ref: No age difference, max. 2 yrs.)
Man is 3-7 yrs older
Man is 8 or more yrs older
Woman is 3 or more yrs older
Age of the woman (Ref: 40-49 yrs old)
30-39 yrs old
Occupational homogeneity (defined by occupational
field) (Ref: Heterogeneous couples, man in
sales/office jobs)
Man is not in employment
Homogeneous – sales/office jobs
Homogeneous – health
Homogeneous – education and teaching
Homogeneous – technical jobs
Homogeneous – other job areas
Heterogeneous – Man in health, education, technical
& other jobs
Age of the youngest child (Ref: No children under 18
yrs)
Under 3 yrs
3-10 yrs
11-17 yrs
Interaction: Age of woman * Age of youngest child
30-39 yrs old * Child under 3 yrs
30-39 yrs old * Child 3-10 yrs
30-39 yrs old * Child 11-17 yrs
Interaction: Age of woman * Age constellation
30-39 yrs old * Man is 3-7 yrs older
30-39 yrs old * Man is 8 or more yrs older
30-39 yrs old * Woman is 3 or more yrs older
1
0,85
0,87
0,81
1
1,03
1
1
0,80
0,73
0,67
1
0,86
1
1
0,97
0,97
1,13
1
1,38***
1
1
1,03
1,23
1,05
1
2,12***
1
1,23
1,56**
0,09***
0,06***
0,18***
0,66
0,89
1,20
1,58**
0,09***
0,06***
0,18***
0,66
0,89
1,50
1,55**
1,36
0,83
1,48
0,88
0,86
1,49
1,56**
1,33
0,83
1,52*
0,89
0,85
1111
0,38***
0,30***
0,46***
0,35***
0,28***
0,42***
0,22***
0,12***
0,24***
0,25***
0,19***
0,27***
1,17
1,13
3,23**
0,71
0,36***
1,41
1,13
1,44
1,98
0,90
0,57*
1,37
Improvement of fit (df):946,22 (38) *** (Model 1)
991,89 (50) *** (Model 2)
Controlled for size of residence, place of residence in East or West Germany, mar-
ried/unmarried cohabitation
Coefficients are significant: p<0.01***; p<0.05**, p<0.1
Case selection: Employed female academics in heterosexual couples, both partners German
citizens, women between 30 and 49 years old, men born after 1943.
Source: Own calculations, German Micro-census 2004, unweighted.
The interaction effect between age of children and women’s birth cohort reveals that
this strategy of reconciling work and family duties through the woman’s reduced work-
ing hours is followed more often by older rather than younger female academics.15

15 This corresponds with the finding that in older cohort, part-time and full-time dual-earner
arrangements were equally likely, whereas in the younger there was a division between
full-time dual-earner or male-breadwinner families.

Page 14

324Rusconi & Solga: Determinants of and obstacles to dual careers in Germany
Younger female academics with 11-to-17-year-old children – who thus had their chil-
dren at a relatively early age compared to older female academics with children of the
same age – have a higher risk of being employed in a non-professional position.16 In
addition, younger female academics with 3-to-10-year-old children have clearly lower
chances of carrying out a full-time professional occupation (than a part-time profes-
sional job) compared to older female academics with children of the same age.17
With regard to age constellation, our results reveal almost no effects (cf. model 1
in Table 6). If women are employed, female academics of the younger cohort have
higher chances to have a full-time professional occupation than a part-time one and
in this regard differences among couples’ age constellations are rather small. Only
in the younger cohort do women with a considerably older partner (8 or more years)
have a lower probability of a full-time rather than part-time professional occupation
compared to women in couples with other age constellations.18
Finally, occupational homogeneity between partners generally enhances women’s
chances not only of being employed, but also of pursuing a professional occupation.
Female academics in homogeneous couples in health and education/teaching have a
higher likelihood of being employed (at least) part-time in a professional occupation
(odds ratios for “no professional jobs” versus part-time professional jobs/reference
category are smaller than 1 and for “full-time professional positions” are not signifi-
cant). Employed female academics in homogeneous couples in technical jobs have
higher chances of working in part-time and full-time professional positions (see also
section 4). Among homogeneous couples, the chances for female professional ca-
reers in sales/clerical occupations seem to be divided because, on the one hand, they
have a higher risk of being employed in non-professional jobs (odds ratio = 1,58),
yet on the other hand, if they do practice a professional occupation, they have a
greater chance of doing so full-time (odds ratio = 1,56). All together, these results
speak in favor of more cooperative (rather than competitive) relationships within
occupationally homogeneous couples.
To conclude, children constrain the opportunities for full-time professional ca-
reers of women in academic couples, and this more so in the younger than the older
cohort. Given the same age of children in 2004, this cohort difference expresses the
women’s age at the transition to motherhood. It indicates that the younger female
academics are at child birth, the lower are their career chances. The couples’ age
constellation plays a minor role in shifting the likelihood of female employment in a
professional position. And finally, occupational homogeneity between partners
seems to enhance the professional careers of women, which suggests that such ho-
mogeneity constitutes shared (borrowed) social capital in couples rather than in-
creasing (hierarchical) competition between the partners.
In our final step, we now examine the importance of the age constellation, occu-
pational homogeneity and children for the realization of dual careers in academic
couples, given that the hurdle of a dual-earner arrangement has already been suc-
cessfully overcome. Accordingly, our multivariate analyses only include couples in

16 Odds ratio difference = 0,86 (n.s.)*3,23 = 2,8
17 Odds ratio difference = 2,12*0,36 = 0,76
18 Odds ratio difference = 1,23 (n.s.)*0,57 = 0,70

Page 15

Zeitschrift für Familienforschung, 19. Jahrh., Heft 3/2007, S. 311-336325
which both partners are employed (Table 7). The results show that children and oc-
cupational homogeneity are the main factors that hinder or increase the chances of
academic couples to pursue dual careers.
Table 7: Multinomial logistic regressions on the chance of realizing dual careers
in academic couples (reference category: no dual career)
(odds ratios, N = 3,284 couples in which both partners are employed)
Dual career: both partner
full-time
Dual career: man full-time
– woman part-time
Mod. 1Mod. 2
1
1,22
1,55**
0,98
1
1,74***
1
Mod. 3
1
1,22
1,52**
0,99
1
1,96***
1
Mod. 1
1
1,14
1,36*
0,94
1
0,91
1
Mod. 2
1
1,08
1,34
0,99
1
0,73
1
Mod. 3
1
1,09
1,34
0,99
1
0,75
1
Age constellation (Ref: No age difference, max. 2 yrs.)1
Man is 3-7 yrs older
Man is 8 or more yrs older
Woman is 3 or more yrs older
Age of the woman (Ref: 40-49 yrs old)
30-39 yrs old
Occupational homogeneity
(Ref: Heterogeneous couple)
Homogeneous couple
Age of the youngest child
(Ref: No children under 18 yrs)
Under 3 yrs
3-10 yrs
11-17 yrs
Interaction: Age of woman * Age of youngest child
30-39 yrs old * Child under 3 yrs
30-39 yrs old * Child 3-10 yrs
30-39 yrs old * Child 11-17 yrs
Interaction: Age of woman * Age constellation
30-39 yrs old * Man is 3-7 yrs older
30-39 yrs old * Man is 8 or more yrs older
30-39 yrs old * Woman is 3 or more yrs older
Interaction: Age of woman*
Occupational homogeneity
30-39 yrs old * Homogeneous couple
1,16
1,29
1,02
1
1,32***
1
5,52***
1
5,59***
1
7,40***
1
3,97***
1
3,94***
1
4,74***
1
0,63***
0,42***
0,56***
0,80
0,63***
0,59***
0,79
0,63***
0,58***
3,40***
4,17***
2,68***
3,82***
3,72***
2,57***
3,80***
3,72***
2,55***
0,68
0,40***
1,17
0, 70
0,41***
1,14
0,96
1,33
0,50
0,97
1,35
0,50
0,92
0,67
1,05
0,92
0,70
1,05
1,12
1,03
0,69
1,11
1,04
0,69
0,57**0,69
Improvement of fit (df): 1002,63 (28) *** (Model 1)
1034,74 (40) *** (Model 2)
1044,00 (44)*** (Model 3)
Controlled for place size, place of residence in East or West Germany, married/unmarried co-
habitation.
Coefficients are significant: p<0.01***; p<0.05**, p<0.1*
Case selection: Both partners employed, heterosexual couples, both partners German citizens,
women between 30 and 49 years old, men born after 1943. Couples with atypical dual-career
constellation were not taken into account due to the small sample size.
Source: Own calculations, German Micro-census 2004, unweighted.
As expected, children reduce the probability of full-time dual careers (indicated by
the significant main effect of children in model 1)19 and particularly the chances of

19 The negative effect of children on full-time dual-careers does not vary among couples
with different age constellation (data not shown). For space reasons, the results of the