Content uploaded by Lynn Prince Cooke
Author content
All content in this area was uploaded by Lynn Prince Cooke on Nov 20, 2014
Content may be subject to copyright.
Household Bargaining and Risk of Divorce
‘Doing’ Gender in Context:
Household Bargaining and Risk of Divorce in Germany and the United States
Running Head: Household Bargaining and Risk of Divorce
Lynn Prince Cooke
Department of Social Policy &
Nuffield College
Oxford University
Oxford OX1 1NF, UK
+44 (0)1865 280 370 phone
+44 (0)1865 270 324 fax
Word count, excluding cover page: 13,633
Earlier versions of these analyses were presented at the International Sociological Association RC28 semi-
annual meeting in Neuchatel, Switzerland, 7-9 May 2004, and the American Sociological Association
annual meeting in Atlanta, 16-19 August 2003. Many thanks are due to Fran Goldscheider for her helpful
comments regarding multiple aspects of this research project.
© 2005, Lynn Prince Cooke. All rights reserved. This paper is for the reader’s personal use only. This
paper may not be quoted, reproduced, distributed, transmitted or retransmitted, performed, displayed,
downloaded, or adapted in any medium for any purpose, including, without limitations, teaching purposes,
without the Author’s express written permission. Permission requests should be directed to
lynn.cooke@nuf.ox.ac.uk
0
Household Bargaining and Risk of Divorce
‘Doing’ Gender in Context:
Household Bargaining and Risk of Divorce in Germany and the United States
Former West German policy reinforced the traditional male breadwinner model
of the husband in employment while his wife performs domestic tasks, whereas in the
United States, policy generally remains silent regarding the private sphere. Explored
here using U.S. and German panel data are effects of different household divisions of
paid and unpaid labor on couples’ risk of divorce. In Germany, any moves away from
the traditional gendered division of labor increase the risk of divorce. In the United
States, the greater divorce risk associated with wives’ increasing earnings is offset by
husbands’ increasing share of domestic tasks until wives’ earnings greatly exceed their
husbands’. In the latter case, U.S. female breadwinner couples where the wife performs
about half of the domestic tasks have much lower log-odds of divorce than traditional
male breadwinning couples in either country. These results indicate that policy shapes
how couples “do” gender in the intimate sphere, with reinforcement of traditional
divisions of labor reducing couple adaptability.
Key words: gendered division of labor, divorce, longitudinal analysis, international
comparison
1
Household Bargaining and Risk of Divorce
‘Doing’ Gender in Context:
Household Bargaining and Risk of Divorce in Germany and the United States
A growing body of evidence indicates that the rules of economic exchange do
not predict who does the housework once wives’ relative earnings exceed their
husbands’. Although the division of housework tends to become more equitable as
wives’ relative household earnings increase from none to about half, it then reverts to a
more traditional division as wives become the primary breadwinner (Bittman et al 2003;
Brines 1994; Greenstein 2000). This latter phenomenon has been attributed to couples
“doing” gender in their marital relationship (Bittman et al 2003; Brines 1994). By doing
gender, people actively manage social interactions in light of normative expectations
(Fenstermaker Berk 1985; West and Zimmerman 1987). The division of housework in
particular reflects the “material embodiment of wifely and husbandly roles, and
derivatively, of womanly and manly conduct” (West and Zimmerman 1987: 144).
Consequently, housework produces both a material and symbolic product of marriage so
that what would seem the fairest division under the rules of exchange does not
necessarily occur within the home (Fenstermaker Berk 1985; see also Hochschild 1989).
The first contribution of this article is to turn attention to “so what?” If the
division of housework reflects couples’ negotiation of intimate life, what is the effect of
these negotiations on marital stability? To date research has focused on the effects of
the division of paid labor on marital stability; little is known about effects resulting from
the division of unpaid labor. Two schools of thought theorize the impact of the
household division of labor on marital stability, and these offer competing hypotheses.
2
Household Bargaining and Risk of Divorce
The first is that gender specialization benefits family solidarity because it increases
couples’ mutual dependence (Becker 1981; Parsons 1942, 1953). Under this
specialization and trading model (Oppenheimer 1997), women’s employment poses a
threat to the benefits of specialization because it reduces women’s economic dependence
upon men, in turn predicting a greater risk of divorce (Becker 1985). By extension,
although never assessed, husbands’ greater domestic participation also threatens the
mutual dependence created by specialization, and so should also increase the risk of
divorce.
The second school of thought evolves from social exchange and bargaining
models (Blau 1960; England and Farkas 1986; McElroy and Horney 1981). Couples
negotiate the division of paid and unpaid labor to a unique equitable distribution within
the family based on relative wages, preferences, etc. These models hold that alternatives
to the marriage are important determinants of relative bargaining power that influence
possible divisions (McElroy and Horney 1981). In these models, women’s employment
and economic independence more generally increase her ability to invoke a credible
threat of divorce if a more favorable division of domestic labor cannot be negotiated.
Under this dynamic, husbands’ greater domestic participation should decrease the risk of
divorce.
To explore which of these competing hypotheses reflects reality, we first assess
whether husbands’ share of unpaid, domestic tasks increases the risk of divorce as would
be predicted by the specialization and trading model, or decreases the risk of divorce as
predicted by social exchange and bargaining models. We then explore the effect of any
3
Household Bargaining and Risk of Divorce
compensatory actions that suggest doing gender in the domestic sphere as women’s
relative earnings rise, which falls outside of either model’s prediction. With the
arguments to date, however, we are left with another set of competing hypotheses. If
doing gender in this way represents a process by which dual-earning couples neutralize
gender deviance (Bittman, et al 2003; Greenstein 2000) and doing so is beneficial to
marital relations, it should decrease the risk of divorce. If, instead, wives taking on a
greater share of domestic tasks as their relative earnings exceed their husbands’ reflects
a display of relative gender power running counter to what is perceived as fair under
notions of distributive justice, it should increase the risk of divorce.
Doing gender, however, is an active process reflecting the institutional shaping
of gender relations. Structural and ideological incompatibilities between the home and
workplace limit women’s ability to achieve equity in either sphere (Ferree 1991;
Hartmann 1981). The degree to which institutional factors reinforce more or less
traditional divisions of labor varies across industrialized societies (Lewis 1992). The
second contribution of this article is to explore how household divisions of labor and any
associated risks of divorce vary in the socio-political context. To do so, we use the U.S.
Panel Study of Income Dynamics and the German SocioEconomic Panel to compare the
risk of divorce among couples as they marry and negotiate the household division of
labor in the United States, where there is less institutional support for the traditional
male breadwinner model, with West Germany, where there has been strong institutional
support for the traditional gendered division of labor. Panel data are used, as they are
most suitable for assessing risk of life transitions such as divorce.
4
Household Bargaining and Risk of Divorce
THE HOUSEHOLD DIVISION OF LABOR
The desirability of the gendered division of labor—when husbands specialize in
economic production while wives specialize in domestic (re)production—is judged
differently depending upon whether one is theorizing about household versus individual
outcomes. At the household level, the specialization and trading model (Oppenheimer
1997) holds that mutual dependence (Becker 1981) and family solidarity (Parsons 1953)
are created when partners specialize then trade the fruits of their specialties. When
women are instead economically independent, marriage is less advantageous to them and
divorce rates are predicted to rise (Becker 1981, 1985).
Empirical evidence of the direct relationship between women’s employment and
marital instability, however, is mixed. Some studies report a positive relationship
(Becker, Landes, and Michael 1977; Brines and Joyner 1999; Ruggles 1997; South
2001), but others find that the positive relationship stems from changes in the pool of
available partners (Aberg 2003; South and Lloyd 1995). Still other research finds no
significant relationship (Bumpass, Martin, and Sweet 1991; Hoffman and Duncan 1995;
Greenstein 1990, 1995) or a relationship only when the marriage is an unhappy one
(Schoen et al 2002). Consequently, it is not clear that women’s participation in the
employment sphere is directly harmful to family stability.
Oppenheimer (1988, 1997) argues the opposite: as male employment has
become increasingly precarious in post-industrial economies, women’s employment is
desirable for family economic flexibility. Sweeney (2002) finds evidence that employed
U.S. women are more likely to marry than unemployed women. Moffitt (2000) finds
5
Household Bargaining and Risk of Divorce
that U.S. marriage rates have been declining at a faster rate among the least-educated
women. In contrast, the greater income associated with more highly-educated, dual-
career couples appears to offset any negative effects of more educated women’s
economic independence (Moffitt 2000). Hoffman and Duncan (1995) report that U.S.
women’s higher wages decrease the likelihood of divorce. This evidence supports
Oppenheimer’s contention that women’s employment in advanced industrial economies
facilitates rather than jeopardizes marriage and marital stability, refuting the desirability
of specialization in the economic sphere.
From an individual perspective, specialization is problematic because the mutual
dependence is not a marker for equality within the marriage (Goldscheider and Waite
1991). As argued within the social exchange or bargaining literature, dominant power is
held by the person who is less dependent on the relationship in terms of having attractive
alternatives (Blau 1960; Emerson 1962; Thibaut and Kelley 1959), with economic
resources a primary source of power (Blood and Wolfe 1960; Weber 1958). Economic
resources are more transferable than an investment in a particular relationship and
children, so a woman’s specialization in the domestic sphere reduces her outside
alternatives to a given marriage (England and Farkas 1986). In addition, having children
by a prior relationship is not advantageous to women seeking a new partner, especially if
she requires that the new partner contribute his own resources to support these children.
Consequently, the gendered division of labor ex ante favors husbands by giving
them direct access to economic resources and superior alternatives to the marriage
(England 1993). This enables them to negotiate solutions more favorable to themselves
6
Household Bargaining and Risk of Divorce
in terms of leisure time and their assistance with the amount or type of domestic tasks
(Blau, Ferber, and Winkler 2002). There is also evidence that husbands expropriate
more of the family economic resources for their own behalf (Blau et al 2002), even those
transfers specifically intended for other family members (Lundberg, Pollack, and Wales
1997). Further, a wife’s economic dependency leaves her vulnerable to a husband’s
exploitation or abuse and, in the case of his death or desertion, poverty.
It is not surprising, therefore, that as female wage rates rose after World War II,
women joined the labor force in increasing numbers. About two-thirds of women aged
15 to 64 across industrialized countries are in the labor force, although this rate varies
from a low of less than 50% in southern European countries to a high of over 75% in
Scandinavian countries (OECD 2000), with married women’s participation rates even
more varied (Jaumotte 2003). Under social exchange and bargaining models, the rising
female employment was expected to lead to a revolution in the gendered division of
domestic labor. In the late 1980s, however, Hochschild (1989) deemed the revolution
“stalled.” Controlling for employment or earnings, time in domestic tasks changes as
women and men move into and out of different familial states, with women increasing
their housework hours when in unions while men decrease theirs (Gupta 1999a; South
and Spitze 1994).
To explain the anomaly, Fenstermaker Berk (1985) argues that the home is a
“gender factory,” producing and reproducing intimate identities of masculinity and
femininity. “Simultaneously, members “do” gender, as they “do” housework and child
care, and what [has] been called the division of labor… is the mechanism by which both
7
Household Bargaining and Risk of Divorce
the material and symbolic products of the household are realized” (Fenstermaker Berk
1985: 201). The home is just one arena in which we do gender, but is the primary site of
our intimate identities and therefore an important one.
Doing Gender and the Division of Labor
The concept of “doing” gender reflects that we are not simply born into a sex or
gender identity or merely functioning in a gender role, but that gender is a “routine,
methodical, and recurring accomplishment” (West and Zimmerman 1987: 126). Social
interactions provide the context for reinforcing the proscribed essentialness of gender.
These interactions do not express natural differences between women and men, but
produce the differences (Goffman 1977). By doing gender, we reflect the social
structure as well as derive relative power consequences of gender category membership
(West and Zimmerman 1987). In this way, as individual members of society, we
actively replicate gender hierarchies in social interactions.
Historically, the division of labor within industrialized societies produces and
reproduces gender hierarchies (Ferree 1990; Hartmann 1981). Early in industrialization,
policies supporting family wages for men and marriage bars for women, along with
protective legislation limiting women’s work activities, locations and hours, all further
reinforced men’s dominance in paid labor (Goldin 1990; Lewis 1992). Despite the
elimination of such explicitly gendered labor force policies across many countries,
gender differences in employment persist, including both horizontal and vertical
segregation and a gender wage gap even after controlling for education and experience
8
Household Bargaining and Risk of Divorce
(Blau et al 2002; Harkness and Waldfogel 1999). One argument is that women’s
continued responsibility for the domestic sphere inhibits their ability to attain
employment equality with men (Ferree 1990; Hartmann 1981; Hobson 1990). So as an
interlocking system, the gendered nature of both paid and unpaid work blocks the ability
to achieve gender equality in either domain (Ferree 1990: 874).
Evidence of this is that, contrary to the “logic of the pocketbook” (Hochschild
1989), as women’s earnings exceed those of men, an even more traditional division of
domestic tasks emerges (Bittman et al 2003; Brines 1994; Fenstermaker Berk 1985;
Greenstein 2000; Hochschild 1989). Using the Panel Study of Income Dynamics,
Brines (1994) finds that the predictions of economic exchange hold for wives, but not
for husbands. As U.S. wives’ relative earnings increase, wives decrease their hours of
domestic tasks in a linear fashion predicted under exchange models. U.S. husbands,
however, decrease their own domestic hours as wives become the primary breadwinner.
She interprets this finding as men’s need for “gender display,” with manhood an
achieved status put under threat when wives take on the traditionally male economic role
in the household.
Greenstein (2000) uses the more detailed housework data from the National
Survey of Families and Households and models the division of domestic tasks with, inter
alia, both absolute hour and proportional measures. Using the absolute hours measure,
9
Household Bargaining and Risk of Divorce
his findings concur with Brines’.1 There is a negative linear relationship between U.S.
wives’ relative earnings and housework hours and a curvilinear relationship between
economic dependence and husbands’ housework hours, indicating that when U.S.
husbands are more dependent upon their wives, their housework hours decline. He
argues that proportional measures are more appropriate, however, as they “are more
likely to capture the distributive justice or equity aspects of the division of housework”
(Greenstein 2000: 325). Using proportional measures, he finds that both husbands and
wives adjust their behavior to more normative gender divisions. So while primary
breadwinning wives may perform fewer total hours of housework because more of these
tasks are provided by the market (i.e., paid domestic help, laundry service, eating in
restaurants) or left undone, these women’s share of domestic tasks is greater than wives
in couples with more economic equality.
Using detailed time diary data, Bittman et al (2003) find that Australian couples
also do gender in terms of compensating behavior in the division of domestic tasks as
wives’ earnings exceed their husbands’, but who engages in the compensating behavior
differs. Australian and U.S. men display a slight curvilinear relationship, but in both
countries, “men’s housework is much more impervious to relative earnings than
women’s is” (Bittman et al 2003: 208). As Australian women’s relative earnings
increase above the point of equality, they respond by increasing their domestic hours.
1 Gupta (1999b) re-analyzes Brines’ data to conclude that the effect is caused by outliers. While this is
technically true, to be a female breadwinning family is to be an outlier, so to analyze their dynamics at the
present time necessarily demands analysis of couples at the extreme.
10
Household Bargaining and Risk of Divorce
Bittman and his colleagues offer that Australian women’s corrective response is larger
than that found for U.S. women because institutional differences make women’s primary
breadwinning more anomalous in Australia than in the United States. The “family
wage” central to strong male breadwinner nation-states was part of Australian
governmental wage setting (O’Connor, Orloff, and Shaver 1999). Consequently, while
doing gender occurs in interactions at the individual level, its patois derives from the
institutional setting (Ferree 1990; West and Zimmerman 1987), with corrective action
more pronounced where more traditional gender roles have been reinforced by policy.
This proposition that policy alters the ways in which couples negotiate the household
division of labor bears elaboration and further testing.
Doing Gender in its Policy Context
Jane Lewis (1992) suggests classifying countries as ranging from “strong” to
“weak” male breadwinner states by the extent to which policy reinforces men’s
preferential access to employment and women’s responsibility for the unpaid care work
in the private sphere. While Australia’s family wage policy is generally indicative of
support for a male breadwinner model, Australia was also the only country to have
comparable worth policies until the mid-1990s. As a result, the Australian gender wage
gap of 88% at that time was among the smallest in industrialized societies, appreciably
smaller than the 78% U.S. gap (Harkness and Waldfogel 1999: 21).
Germany is the ideal-typical strong male breadwinner state (Lewis 1992). At the
end of World War II, West Germany founded a new political system based on “natural
11
Household Bargaining and Risk of Divorce
law” (Naturrecht), stemming from a “pre-political” patriarchal order ordained by God
(Moeller 1993). Strengthening the patriarchal family as an institution dominated West
German federal policy under Konrad Adenauer during the 1950s and 1960s, with social
provisions favoring male breadwinners with non-working wives (Gerhard 1992; Ostner
1992; Zimmerman 1993). Income splitting for taxation purposes, particularly beneficial
to high-income, single-earner families, was introduced in 1958. In 1961, on a platform
of “yes to family,” mothers were deemed the only satisfactory educators of their children
(von Oertzen 1999). Schools were subsequently set up to finish after two hours on one
day, six the next, and are closed over the lunch hour (Ostner 1993). Until 1977,
domestic responsibilities continued to be recognized as West German women’s legal
duty (Hantrais 1994).
West German women’s educational attainment and employment lagged during
this time. Fewer West German women than men graduated from the highest secondary
school tracks and went on to university, and there are clear gender differences in the
type of occupational training selected (Geschka 1991). A 1966 Federal report
highlighted gender disparities in educational attainment and concluded that women were
failing to exploit fully their right of education—not to the detriment of their employment
prospects, but to the detriment of the education of the next generation (von Oertzen
1999). At the time of economic unification with East Germany in 1990, only 44 percent
of West German married women were employed, and only half of this percentage were
employed full-time (Ostner 1993).
12
Household Bargaining and Risk of Divorce
In contrast, U.S. policy since World War II has primarily reflected the liberal
tradition and addressed women’s ability to compete in the labor market.2 Employment,
training and education discrimination on the basis of gender became illegal under Title
VII of the Civil Rights Act of 1964. During the 1970s, laws expanded enforcement of
this legislation, promoting equal educational opportunities and job training. U.S.
women’s secondary completion rates have historically been higher than their male
counterparts. Since 1982, more bachelor’s degrees have been conferred on U.S. females
than males, women earn more associate and master’s degrees than men, and are coming
close to parity in first-professional and doctoral degrees (U.S. National Center for
Education Statistics 2000). These policies also encourage U.S. women’s labor force
participation, with two-thirds of U.S. married women with children aged six to 17
employed, as are almost 60% of married women with children under the age of six (U.S.
Bureau of Labor Statistics 2004). Unlike West German policy, U.S. policy remains
silent regarding who is responsible for the private sphere (Leibfried and Ostner 1991).
This is not to say that gender equality is a given with liberal market economies (see
Hartmann 1981), just that a gendered division of domestic labor is not specifically
reinforced by U.S. policy.
From these divergent policy paths, the institutional framework defining where
gender should be “done” vis-à-vis the division of labor varies across the countries.
2 Feminist or reformer concerns for women as mothers also play a key role in U.S. policy, particularly at
the end of the 19th and beginning of the 20th century. Orloff (1996) provides a review of the scholarship
documenting this era.
13
Household Bargaining and Risk of Divorce
Because of this, West Germans should express greater support for the traditional
gendered division of labor than U.S. persons. Figure 1 displays the percentage of
representative samples of West German and U.S. women and men surveyed for the 1994
International Social Survey Program3 that agree or disagree with the statement, “A
man’s job is to earn money; a woman’s job is to look after the home and family.” In
both countries, more people now disagree than agree with that statement, but support for
the traditional male breadwinner model is appreciably stronger in West Germany.
Almost 40% of West German men still agree with the statement, as compared with just
one-quarter of U.S. men. Women within their countries report more non-traditional
attitudes than men, but over one-third of West German women agree with the traditional
division, as compared with less than 20% of U.S. women. Analyzed here is the extent to
which these more traditional attitudes translate into more traditional household divisions
of labor within West German as compared with U.S. couples, and whether different
divisions of labor alter the risk of divorce for couples in either country.
THE DIVISION OF LABOR AND MARITAL STABILITY
The specialization and trading model relates women’s rising employment to
rising risk of divorce, but has not explicitly predicted effects of husbands’ greater
3 More information about the International Social Survey Program can be found at http://www.issp.org.
The organization has fielded a more recent survey on this topic in 2002, but the 1994 wave is selected as it
falls within the observation window of the larger analysis here. Relative country means also remain
similar across the two waves.
14
Household Bargaining and Risk of Divorce
domestic contribution. If specialization creates an essential mutual dependence, it can
be deduced within this framework that men’s increasing domestic contribution
undermines specialization and so would also be predicted to increase the risk of divorce.
In contrast, social exchange (Blau 1960), contract (England and Farkas 1986),
and game theoretic bargaining models (Manser and Brown 1980; McElroy and Horney
1981) view the observed household division of labor as the result of couple negotiations
that reflect each person’s relative power and resources. Divorce is not the given outcome
but forms the lower bound of an acceptable outcome, a person’s threat point, or BATNA
(Best Alternative to a Negotiated Agreement). If women find themselves unable to
negotiate a favorable division of domestic tasks, economically independent women are
able to then leave the marriage. As more women gain economic independence, more are
able to threaten divorce.
This suggests that the higher divorce rates since World War II are not reflecting
just effects of women’s rising employment; they reflect men’s resistance to changing
their domestic behavior in response to women’s rising employment. Lennon and
Rosenfeld (1994) find that women with more alternatives to the marriage perceive
unequal divisions of domestic tasks as unfair, whereas women with fewer alternatives
report greater acceptance of the situation. Schoen and his colleagues (2002) find that
U.S. women’s employment only increases the risk of divorce when the marriage is an
unhappy one. More than two decades ago, Huber and Spitze (1980) found that while
wives’ thoughts of divorce increase with their own employment, they decrease with
husbands’ increasing housework contribution. To date, however, analyses of the effect
15
Household Bargaining and Risk of Divorce
of wives’ employment on divorce have not controlled for possible countervailing effects
of husbands’ domestic contribution. If the social exchange and bargaining predictions
are correct, husbands’ greater share of domestic tasks should decrease the risk of
divorce.
It should be noted that Lundberg and Pollak (1993, 1994) question whether
divorce is the true threat point in family bargaining. They argue that even if couples
cannot reach agreement, there are still benefits such as shared living economies and
enjoyment of the children that make divorce a less desirable alternative. They suggest
that failing agreement, couples first revert to a non-cooperative strategy reflecting the
traditional roles under the gendered division of labor. The absence of cooperation in
marriage, however, results in either a poorly-maintained home or fewer children, since a
woman will only produce what she can manage on her own (Lundberg and Pollak 1993).
There is evidence of these dynamics across countries. Couples in which the husband
contributes less to housework or childcare have a lower risk of second children in
Germany (Cooke 2004), Hungary and Sweden (Oláh 2003), and the United States (Torr
and Short 2004). Consequently, only the effects of the division of housework on
likelihood of divorce shall be analyzed here.
Neither the specialization and trading nor social exchange and bargaining models
offer predictions for the effects of compensating behavior on the risk of divorce. For
this we link the hypothesized reasons for doing gender in the household division of labor
to possible family effects. Bittman et al (2003) and Greenstein (2000) argue that the
compensating behavior in the form of women’s greater domestic share when she is also
16
Household Bargaining and Risk of Divorce
the primary breadwinner neutralizes gender deviance in a marriage. If minimizing this
deviance is essential not only to intimate identities but to relationship stability, doing
gender in this way should decrease the risk of divorce. If, instead, the rules of exchange
and distributive justice dominate in successful couples, female breadwinners
compensating by taking on a greater share of domestic tasks might be performing a stop-
gap measure within an inherently unfair situation. If so, the risk of divorce should be
higher in couples exhibiting the compensating division of domestic tasks than in couples
where there is gender equity. By gender equity we refer to when wives’ relative
contribution to earnings is the same as husbands’ relative contribution to housework,
ranging from 0, where wives contribute nothing to earnings and husbands contribute
nothing to housework, to 100, indicating women are the breadwinners while husbands
assume all domestic tasks.
The effects of the household divisions of paid and unpaid labor on couples’ risk
of divorce are compared in the United States and the former West Germany to see
whether they vary as policy support for the male breadwinner model has varied. For this
to be a meaningful comparison, the propensity to divorce in each country must be
similar, so we observe differences in effects, not differences in overall likelihood of
divorce. The current incidence of divorce is quite similar in the two countries, at 49.1
divorces per 100 marriages in the United States (U.S. Census Bureau 2000) as compared
with 46.0 divorces per 100 marriages in unified Germany (EUROSTAT 2000). As
illustrated in Figure 2, attitudes toward marriage and divorce are also highly similar in
the two countries. The figure displays the mean reports by country for three questions
17
Household Bargaining and Risk of Divorce
from the 1994 International Social Survey Program that assess attitudes about marriage
and divorce: 1) married people are generally happier than single people, 2) a bad
marriage is better than being single, and 3) divorce is the best solution to a bad marriage.
In both countries, the samples tend not to agree that married people are happier, strongly
disagree that a bad marriage is preferable to being single, and moderately agree that
divorce is the best solution to a bad marriage, with West Germans slightly more likely to
agree than U.S. respondents.
DATA
Wives’ relative earnings, work hours, husbands’ share of domestic tasks, number
of children, and the risk of divorce vary across the marital life course. The most suitable
way to assess these dynamic relationships is with event history analysis (Allison 1984;
Yamaguchi 1991), which requires longitudinal data. The German SocioEconomic Panel
(GSOEP) and the Panel Study of Income Dynamics (PSID) are selected for the analyses.
The GSOEP is a longitudinal study of private German households where all
household members over the age of 16 are interviewed annually for data on the
preceding year. The first wave occurred in 1984 with a representative sample of 12,290
people in 5,921 households in the former West Germany. In June 1990, sampling
extended into the former East, but East Germans shall not be included in this analysis as
the male breadwinner model was not institutionally reinforced in that region. The
constitution adopted by the East in 1949 enforced women’s right and obligation to work,
supported by extensive policy provisions enabling women to combine work and having
18
Household Bargaining and Risk of Divorce
children (Moeller 1993; Ostner 1993; Zimmerman 1993). After unification, economic
constraints dramatically increased unemployment in the East and made East Germans
less likely to make any family transition—into or out of marriage (Münz and Ulrich
1995; Witte and Wagner 1995). Too few East couples in the sample married (69) and
divorced (21) during the observation window to conduct separate East analyses.
The PSID is a longitudinal study of U.S. individuals and the family units in
which they reside, beginning in 1968 with a representative sample of 4,800 families in
which the head only is normally interviewed. Although two thousand Latino households
were added to the panel in 1990, they were subsequently dropped in 1995 and a new,
much smaller immigrant sample added. Consequently, the only ethnic differentiation
possible in the analysis here is when respondents are Black. Historically, the rate of
marital dissolution has been greater for Black couples (Hoffman and Duncan 1995;
Ruggles 1997).
From each dataset, we select couples marrying for the first time between 1985
and 1995 for which there is at least one year of data following the marriage. These
couples are followed through 1997 in the United States and 2000 in Germany.4 The
year 1985 is selected for the beginning of the observation window, as it is the second
year of GSOEP data collection. The same year is selected for the PSID data in order to
follow couples in both countries during the same historical time period. Couples already
4 The PSID changed in 1997 to be conducted bi-annually and the core sample was reduced by almost 30%.
This created substantial missing data in the 1999 and between waves, so it was decided to end the U.S.
observation window in 1997.
19
Household Bargaining and Risk of Divorce
married as of 1985 are excluded from analysis, since their inclusion biases results with
marriages of longer duration and would not provide an accurate portrayal of family
dynamics affecting the risk of divorce for all couples from the beginning of their
marriage. For example, in Greenstein’s sample drawn from the National Survey of
Families and Households, the average marriage duration was 17 years (2000: 327),
whereas in 1990, the median U.S. first marriage duration was approximately 8 years
(National Center for Health Statistics 1995).
In both panels, new members marrying into the sample can be assigned a weight
of zero if they are not within the original sampling frame. For this reason, unweighted
data are used for the analysis, although a comparison of weighted and unweighted
sample descriptive statistics indicates that the two are highly similar. As a result, the
analytical samples appear representative of German and U.S. couples marrying during
the time period.
The West German sample includes 522 couples, yielding an analytic sample of
4,483 couple-years, reduced to 3,524 due to listwise deletion of missing data.
Subsequent analysis using indicator variables for when data are missing indicates
robustness of key effects. The initial U.S. sample is comprised of 1,122 Caucasian and
368 Black couples, yielding an analytic sample of 9,633 U.S. couple years, reduced to
4,204 due to listwise deletion of missing data primarily caused by attrition from the
panel. When indicator variables are included for the missing data, key U.S. effects
remain robust, although the significance levels of some control variables change.
Further, Lillard and Panis (1994) report that biases from sample attrition in the PSID
20
Household Bargaining and Risk of Divorce
when analyzing marital dissolution are generally mild. Also, demographics and the
division of labor derived from the PSID sample used here have been compared with
cross-sections of other more recently fielded surveys (National Survey of Families and
Households and the International Social Survey Program), and appear similarly
representative.
In the dataset, each year of a couple’s marriage is a distinct observation,
beginning with the first year of marriage and concluding with either divorce or
separation (which are not distinguished in this analysis given the variation in required
waiting periods between separation and divorce), or the final observation year in the
panel. Constructing couple-years in this way automatically incorporates the time-varying
aspects of the independent variables, but also violates the assumption that error terms
not be correlated. Consequently, robust standard errors clustering on a unique couple
identification number are used.
Variables
Table 1 contains the descriptive statistics for the variables used in the analyses.
These statistics are based on the couple-year files constructed for the event history
analyses, so the values of the time-varying variables represent averages over the
observed years of marriage, not a snapshot of couples in any given year of marriage.
Dependent variable. – The dependent variable is a binary variable indicating
whether a couple reports a divorce or separation in a given year. Once a divorce occurs,
the couple is removed from analysis as they are no longer at risk of divorce. During the
21
Household Bargaining and Risk of Divorce
time period, 201 West German couples reported divorcing and 223 U.S. couples (153
White and 90 Black) reported divorcing. These aggregate figures suggest the divorce
rate in Germany is higher than in the United States, when as noted earlier it is roughly
similar in the two countries. The U.S. couples, however, are being followed for three
years less than their German counterparts. As the event history models control for the
effect of time on the risk of divorce, it will be possible to see whether the shorter U.S.
observation window is the reason the observed U.S. sample divorce rates appear lower.
Independent variables. Women’s Employment. In both panels, participants are
interviewed in a given year to ascertain information about their lives over the past 12
months. To ensure that causes of divorce are differentiated from effects, values of the
time-varying independent variables are lagged by one year. For example, if a woman is
in the process of establishing her own household, her share of household earnings would
rise, leading to the erroneous conclusion that her greater earnings caused the transition
rather than resulted from it. Similarly, total household income would decrease as dual-
earner couples become single heads of household. In a troubled marriage, both partners
may also reduce their time spent in household tasks as commitment to the marital home
declines.5
5 Comparing models using lagged versus unlagged independent variables, these sorts of differences are
born out (results available from the author). The substantive effect of U.S. wives’ relative earnings is
larger when using unlagged variables, the effect of U.S. husbands’ share of housework is smaller and
becomes statistically insignificant, and the prophylactic effect of total household income is much larger in
both countries.
22
Household Bargaining and Risk of Divorce
Most analyses of women’s employment on the likelihood of divorce assume that
effects are linear by using a single continuous measure of women’s hours of
employment or wages. It is possible, however, that being a housewife is a
fundamentally different state than being a part- or full-time working wife, particularly
when comparing two countries with varying support for the traditional division of paid
labor. The traditional family might also, in turn, carry different risks of divorce. To
assess these effects, a binary variable is included for when the wife is out of the labor
force, against a referent of wives who are employed.6 If specialization is optimal for
marital stability, these traditional couples should have a lower risk of divorce than dual-
earner couples. Further, as the male breadwinner model has been reinforced more in
Germany than the United States, there should be more of these couples in Germany. As
displayed in Table 1, over half of West German wives are out of the labor force at some
time during the observed years of marriage, as compared with less than 10% of U.S.
wives.
To keep the metrics for the division of paid and unpaid labor commensurate, a
wife’s contribution to the family labor earnings is measured as a percentage (0 to 100) of
the combined labor earnings of the wife and husband. This measure is perfectly
correlated with the relative dependency measure developed by Sorenson and
McLanahan (1987) used in other analyses predicting the division of housework (Brines
6 Doing so proves important to ascertaining true effects, for when not differentiating for when women are
out of the labor force, the effect of women’s relative earnings appears much smaller because it is
camouflaging the higher risk of divorce among U.S. couples where the wife is out of the labor force.
23
Household Bargaining and Risk of Divorce
1994; Greenstein 2000). As shown in Table 1, across all couples, the average household
earnings contribution of U.S. wives is almost twice that of West German wives: 32
versus 18%, respectively.
Wives’ hours of paid work are included to control for the time demands
employment places on her that are theorized to alter the division of housework
separately from the effects of her relative earnings (Blood and Wolfe 1990; South and
Spitze 1994). Other things being equal, the more hours worked in the market, the fewer
available for housework. Time constraints have proven important in predicting second
births among German couples, an effect in addition to women’s earnings’ effects (Cooke
2004).7
Division of Housework. The German data have variables for each partner’s
reported weekly domestic hours, including housework, running errands, yard work and
repairs, and childcare, both during the week and the average hours on Saturday and on
Sunday. The measure of housework used here is calculated as the time spent in all tasks
with the exception of childcare throughout the week. Childcare is excluded as there is
no similar measure within the PSID, but Cooke (2004) finds that German fathers’
relative contribution to childcare proves insignificant in altering the risk of divorce.
The PSID measure of housework is more limited in two ways. First, in contrast
to the GSOEP, the PSID normally relies on a single primary adult—usually the male
adult head if there is one—to provide information for all family members. Most
7 The models were run with and without women’s work hours with no substantive or significant changes
in other effects or the fit of the model, but were left in for the theoretical reasons stated above.
24
Household Bargaining and Risk of Divorce
evidence indicates that respondents overestimate their own housework time and
underestimate the time contribution of others (Shelton and John 1996), with Press and
Townsley (1998) finding that husbands’ reports are less accurate than wives’. The
extent of the possible U.S. husband-only reporting bias was analyzed by comparing a
cross-section of the longitudinal sample with a sample from one of the few PSID dual-
respondent surveys (1985; results available from the author). That analysis indicates
that while U.S. husbands do tend to underestimate their wives’ housework hours, the
relative division of housework is fairly represented in the data once controlling for
husbands’ reports of wives’ estimated weekly housework hours, so there are no
significant biases introduced in the model used here.
The second limitation of the PSID is that it contains a single question asking the
respondent to estimate how many hours are spent in “housework, excluding childcare”
each week. These data were used in the analyses by both Brines (1994) and Gupta
(1999b) to reveal that at the extreme of a husband’s dependence on his wife, there is
evidence of gender display. A single measure does not provide the detail of the German
data, nor is it as rich as the series of questions asked of U.S. respondents on the National
Survey of Families and Households and analyzed by Greenstein (2000).8 None of these
8 Although the National Survey of Families and Households collects more detailed information on
household tasks for each family member, it has only conducted three waves several years apart (1987-88,
1992-94 and 2001-2002), which makes it less suitable than the PSID for conducting event history analyses
of divorce.
25
Household Bargaining and Risk of Divorce
are as precise as the time diary data used by Bittman et al (2003) in their analysis of
Australian couples. But despite the wide range in the crudeness of the housework
measure, results are remarkably consistent in terms of the extent of equity or
compensatory behavior made evident with them. Consequently, concern for the quality
of the housework measure appears more philosophical than applicable to the analysis.
Husband’s relative domestic participation is measured by dividing his weekly
hours in housework by the combined household hours of the wife and husband, yielding
his percentage share of domestic tasks ranging from 0 to 100. Also included is the
square of this term to test for nonlinearity. If the squared term is positive and
significant, this indicates that husband’s greater relative housework contribution at some
point begins to increase the risk of divorce. A relative rather than absolute housework
measure is used because of Greenstein’s (2000) evidence that compensating behavior is
more evident when controlling for the relative rather than the absolute housework
contribution of each. As also noted by Greenstein (2000), perceptions of fairness and
distributive justice are related to the relative, not total amount.
Women’s reported housework hours are included to control for the husband-only
reporting bias in the PSID, as well as to control for when men’s share of housework is
greater because women’s own housework hours decline when they are employed
(Bianchi et al 2000; Goldscheider and Waite 1991; Shelton and John 1996). Including
this measure does not create a problem of multicollinearity with women’s hours of
employment, however, because while employed women’s housework hours adjust to
26
Household Bargaining and Risk of Divorce
reflect competing time demands, the sharpest drop in housework hours since the 1960s
has been among women who are out of the labor force (Gershuny 2000).9
As evident in Table 1, German wives spend more hours in housework on average
than their U.S. counterparts (29 and 17 hours, respectively), although this could be due
to the higher proportion of housewives in the German sample. Overall, the gendered
division of housework appears roughly similar in the two countries, with husbands in
each country performing about one-third of the housework as has been found in other
cross-national studies (Gershuny 2000). Husbands’ share of housework when wives are
employed is less similar in the two countries, with U.S. husbands claiming to perform
44% as compared with German husbands performing 38%.10
The above measures enable us to assess the economic exchange relationship
between the household division of labor and risk of divorce predicted by bargaining
models, but not whether compensatory behavior in the division of domestic tasks
reduces the risk of divorce when the wife’s earnings exceed her husband’s. To assess
this, an interaction indicator variable is created for when wives’ relative earnings exceed
50% and husbands’ share of housework does not exceed 50%. In Germany, only 13%
of wives earn more than their husbands, compared with 24% of U.S. wives. As can be
seen in Table 1, roughly half of these German female-breadwinning couples but just
9 Including women’s housework hours significantly improves the fit of the model, but does not
substantively or significantly alter effects of other variables.
10 Disability or poor health can also alter the household division of labor, but only one or two of these
younger first-married couples report that one of the partners is in ill-health or disabled in any given year.
27
Household Bargaining and Risk of Divorce
one-third of U.S. female-breadwinning couples compensate for wives’ higher earnings
by doing gender in this way (7% of the German and 8% of the U.S. samples). This is
consistent with Bittman et al’s (2003) Australian evidence that compensating behavior
in the domestic sphere is more common in strong male breadwinner states.
Control variables. – Controls for a wife’s age, number of children, total family
income, home ownership and college or university attainment are included in the
models. According to models of assortative mating, people who are older at marriage
are less likely to divorce because of a decrease in possible future marital partners
(Becker et al 1977). But culturally-defined bargaining disadvantages specific to women
at older ages also exist. In many industrialized societies, youth and beauty prove valued
characteristics of women and less important for the marriage prospects of men (England
and Farkas 1986; Parsons 1942). Youth by definition disappears with age, so women
lose their bargaining advantage over time regardless of any other factors. A variable is
included here to control for women’s age. In both country samples, wives are, on
average, 30 years old.
Persons with higher-valued characteristics such as university education or wealth
gain more from marriage and are therefore less likely to divorce net of other factors
(Becker et al 1977). The log of total household income is included to control for wealth
of the family. Men’s education level is positively associated with their domestic
participation (South and Spitze 1994), whereas other studies find no association (Kamo
1991; McAllister 1990), or that the effect disappears once gender ideology is included
(Kamo 1994). Women’s greater educational attainment is associated with less time in
28
Household Bargaining and Risk of Divorce
domestic tasks (Blair and Lichter 1991; Brines 1993; South and Spitze 1994), and is
normally interpreted as an education effect on ideology. There is some U.S. evidence
that the historical effect of wife’s greater educational attainment lowering the risk of
divorce is attenuating in younger cohorts (South 2001). A binary variable is created for
women or men with college or university education, against a referent of less than
college.
In neoclassical economic models of marriage, children and home ownership
represent accrued “marital goods,” so they predict a lower risk of divorce (Becker 1981).
One might also argue that more stable couples are more willing to purchase a home
together or have additional children, suggesting these might reflect a selection bias.
Number of children in the family in a given year is entered into the models as a
continuous variable. Home ownership is measured with a binary variable indicating
when a couple owns their home, against a referent of renting.
The risk of divorce can change as a function of time, irrespective of the
independent variables, so a variable is included for years since marriage. This also
enables assessment of whether the observed country differences in the incidence of
divorce relative to sample size reported earlier are a function of time. A piecewise
constant model including a series of year binary variables representing two to three,
three to four, four to five, and five or more years from year of marriage was also tried,
but any significant time effects in the respective countries proved to be essentially linear
during the observed years of marriage.
29
Household Bargaining and Risk of Divorce
In the U.S. model, a binary variable is also included indicating when the husband
is Black, against a referent of when the husband is Caucasian. Racial marital homogamy
is high: 95% of Black women are married to Black men and 99% of Caucasian women
are married to Caucasian men.
FINDINGS: THE DIVISION OF LABOR AND RISK OF DIVORCE
The changes in the log-odds that a West German or U.S. couple will divorce are
presented in Table 2. Discussion focuses on effects of the variables measuring the
household division of labor. Two models are presented: one modeling the effects of the
household division of labor using continuous measures, and the second adding the
indicator term for when husbands perform half or less of domestic tasks when wives’
earn more than half of the couples’ labor income.
Effects of the Household Division of Labor
Results for Model 1 indicate specialization is optimal for marital stability in
West Germany, but not in the United States. While male breadwinner couples in both
countries are more likely to divorce, the effect only reaches statistical significance in the
United States. Women’s rising relative earnings also predict a rising risk of divorce,
with the effect twice the magnitude in the United States as in Germany. Each percentage
point increase in a wife’s earnings as a percentage of her and her husband’s total labor
earnings increases the log-odds of divorce by 1% in West Germany and 2% in the
United States. Together these results suggest that the male breadwinner couples
30
Household Bargaining and Risk of Divorce
reinforced by policy are the most stable in West Germany, whereas dual-earner couples
are the most stable in the United States provided a woman’s earnings do not exceed her
husbands’. In the United States, being a male breadwinner couple increases the log-odds
of divorce by 0.83, comparable to the increase in risk when a wife is earning more than
41% of the family’s labor income (41*0.02 wife’s earnings effect). So reliance on a
primary breadwinner of either gender among U.S. couples proves more precarious than
when there are two more equal earners in the family.
These relative employment effects, however, do not take into account possible
countervailing effects of the division of housework, for which there are also marked
country differences. German husbands’ increasing share of housework linearly increases
the risk of divorce across the entire range (i.e., the quadratic term is insignificant). In
contrast, U.S. husband’s increasing housework contribution significantly decreases the
risk of divorce to more than offset the effect of wives’ earnings. But the U.S. quadratic
term is also significant, so once U.S. husbands’ housework and wife’s earnings
contributions exceed about 30%, the risk of divorce again begins to increase. Yet the net
effect of husband’s housework contributions equaling wives’ relative earnings still
predicts lower log-odds of divorce until both exceed 83%, which is an extremely non-
traditional division of paid and unpaid labor.
These effects are depicted in Figure 3, which plots changes in the log-odds of
divorce for percentage point changes in wives’ relative earnings matched by identical
changes in husbands’ share of housework (i.e., when wives’ contribute 10% of earnings,
husbands’ share of housework is also 10%; when wives’ contribute half to earnings,
31
Household Bargaining and Risk of Divorce
husbands are doing 50% of the housework, etc.), controlling for other significant effects
related to the division of labor. In other words, it plots changes in the predicted risk of
divorce when couples negotiate a division of paid and unpaid labor reflecting
distributive justice. Only when comparing male breadwinner couples with a traditional
gendered specialization in paid and unpaid labor is the relative risk of divorce lower in
West Germany than in the United States. All other more egalitarian divisions of paid
and unpaid labor increase the risk of divorce in West Germany whereas they reduce the
risk of divorce in the United States.
The second model in Table 2 displays effects of compensatory domestic behavior
attributed to couples doing gender to neutralize gender deviance when wives become the
primary breadwinner (Bittman et al 2003; Brines 1994). For the German couples, when
wives perform a greater share of housework as their relative earnings exceed 50%, the
negative effects associated with her greater relative earnings remain unaltered. For U.S.
couples, this compensatory behavior at the extreme non-traditional division of paid labor
predicts a much lower risk of divorce. This suggests that there are positive effects for
gender equity in the U.S. household division of labor until parity is reached, but
neutralizing gender deviance has a prophylactic effect among those few women who are
primary breadwinners. Yet even under this compensating scenario, the remaining
significance of the effect for husbands’ housework share indicates that these
32
Household Bargaining and Risk of Divorce
non-traditional couples are most stable when husbands perform some share of the
housework.11
In summary, German traditional male breadwinner families are predicted to be
the most stable in that country. Any other divisions of paid and unpaid labor, even
compensatory ones wherein a female breadwinner would retain responsibility for
domestic tasks, increases the risk of divorce. In contrast, the rules of distributive justice
appear to dominate among these first-married U.S. couples until relative earnings
approach a non-traditional extreme. When wives’ relative earnings exceed their
husbands’, couples can reduce the risk of divorce if wives continue to perform a share of
the housework. Further, when controlling for other significant effects related to the
division of labor, U.S. female breadwinner families where the woman does half the
housework have a lower risk of divorce than German male breadwinner couples.12
11 From coefficients in Table 2, Model 2: U.S. female breadwinner, husband contributes half to domestic
tasks = [(100*0.02 wife’s earnings) + (-0.04*50 husband’s share of domestic) + (-0.03*13.50 employed
wife’s mean housework hours) + -2.25 wives’ compensatory domestic share effect in that the husband
does no more than half] = -2.66. Female breadwinner husband does nothing = [(100*0.02 wife’s
earnings) + (-0.03*13.50 employed wife’s mean housework hours) + -2.25 interaction effect that the
husband does no more than half of domestic tasks] = -0.66.
12 From coefficients in Table 2, Model 2: Being a German male breadwinner couples predicts no net
change in the risk of divorce. Computing for being a U.S. female breadwinner couple where the husband
contributes half to domestic tasks yields a substantial decrease in the log-odds of divorce ([(100*0.02
wife’s earnings) + (-0.04*50 husband’s share of domestic) + (-0.03*13.50 employed wife’s mean
housework hours) + -2.25 interaction effect that the husband does no more than half of domestic tasks] =
33
Household Bargaining and Risk of Divorce
Other Household Bargaining Effects
The primary interest here is to compare effects of divisions of paid and unpaid
labor on the risk of divorce in two countries with varying levels of institutional support
for a male breadwinner model. These institutional differences, however, can manifest
indirectly as well, reflecting other gender differences in relative bargaining power. As
noted earlier, if wives have not accrued work experience because they have been reliant
on a male breadwinner, having children increases their dependence and inability to leave
the marriage. In West Germany, where the majority of households are male
breadwinner households, the presence of children significantly reduces the risk of
divorce. The effect is linear, so each additional German child further reduces the risk of
divorce. In contrast, in the United States where the vast majority of couples have two
earners so the vast majority of wives have some accrued work experience, the number of
children does not alter the risk of divorce, a result also found by others (South 2001;
Waite and Lillard 1991).
Another difference in effects that might stem from the difference in the degree to
which policy encourages couples’ economic interdependence versus wives’ dependence
is that of home ownership. In the United States, the interest paid on home mortgages or
loans secured with one’s home is tax deductible. This provides U.S. married women
-2.66). In female breadwinner couples where the husband does no domestic tasks, the predicted net
change in the risk of divorce is still less than zero ([(100*0.02 wife’s earnings) + (-0.03*13.50 employed
wife’s mean housework hours) + -2.25 interaction term] = -0.66).
34
Household Bargaining and Risk of Divorce
with an incentive to join the labor force, enabling the family to afford a larger home in a
better neighborhood, which enhances a husband’s status as well. There are some
deductions related to home ownership in Germany, but they are capped to a very modest
level. As evident in Table 2, home ownership reduces the log-odds of divorce in both
countries. The effect, however, is more than five times as great in the United States as
in Germany (-1.23 versus -0.23, respectively) and is only statistically significant in the
United States for these young couples. While other differences, such as the desirability
of the rental sectors in the two countries, might account for part of this difference, it is
also possible that the economic interdependence within couples created by the U.S.
home mortgage tax provisions is a significant factor. Future research might explore the
effect of policies promoting greater couple economic interdependence, as an antidote to
the negative effects found here of policy reinforcement of the traditional male
breadwinner model and women’s economic dependence.
DISCUSSION: DOING GENDER IN CONTEXT AND ITS EFFECTS ON
MARITAL STABILITY
How we “do” gender reflects historically-reinforced hierarchies shaped by policy
and other institutional factors. Consequently, gender relations vary across socio-political
contexts, with the hierarchy evident in and replicated by the gendered division of labor
in society (Hartmann 1981; Lewis 1992). Only recently, however, have policy
influences and resultant divisions been compared across societies (Baxter 1997; Bittman
et al. 2003; Davis and Greenstein 2004; Fuwa 2004), and none of these analyses
35
Household Bargaining and Risk of Divorce
compared effects of varying household divisions of paid and unpaid labor on marital
stability. The specialization and trading model claims the traditional gendered division
of labor reinforced in some socio-political contexts is optimal for marital stability.
Social exchange and bargaining models contend that women’s rising economic equality
with men alters the credibility of threat points such as divorce through which more
equitable distributions of household labor can be negotiated. Under these models,
equitable divisions of household paid and unpaid labor reflect distributive justice that
should enhance marital stability.
Here, the household division of labor and its effects on marital stability of
couples first married between 1985 and 1995 are compared in the United States and the
former West Germany. U.S. liberal policy encourages female labor force participation
while remaining silent on the private sphere, whereas West Germany implemented
policies after World War II reinforcing women’s legal responsibility for the home and
economic dependence on a male breadwinner. With this historical reinforcement of the
gendered division of labor, more West German couples report the wife is out of the labor
force as compared with U.S. couples, where dual-earner couples are the norm. When a
West German wife is employed, she performs more domestic tasks, displaying more
compensating domestic behavior for her non-traditional economic role than U.S. wives.
This is similar to evidence for couples in Australia, another strong male breadwinner
country as compared with the United States (Bittman et al 2003), and the variation in
individual effects across countries found by Fuwa (2004). Together these findings
highlight that it is not sufficient to look at individual resources in making predictions
36
Household Bargaining and Risk of Divorce
regarding the household division of labor; we must situate effects within the institutional
setting, particularly the extent to which policy reinforces the gendered division of labor.
This determines the extent to which a gendered division of labor is done within the
intimate sphere of the home.
As a counter-argument, one might assert that other cultural differences explain
both the traditional policies and effects found for West Germany as compared with the
United States. Yet the former East Germany shares a common cultural past with West
Germany while the people were socialized for two generations under divergent policies
affecting the division of labor. After World War II, East Germany adopted a Stalinist
constitution that enforced women’s obligation to work (Moeller 1993). To support
maternal employment, East Germany passed the 1950 Mother and Child Care and
Women’s Rights Acts, establishing a network of public childcare centers, kindergartens
and facilities for free school meals, maternity leave, and days off to tend sick children
(Ostner 1993; Zimmerman 1993). The state also mandated developing women’s skill
credentials through education and vocational training, and a larger proportion of East
German women attended professional colleges and university than in West Germany
(Budde 1999). Recent evidence indicates that even after economic reunification, the
division of housework is significantly more egalitarian in the former East than in West
Germany (Cooke 2004), and is similar to the division reported here for U.S. couples.
These differences within Germany suggest it is not just culture reflected in the divergent
household divisions of labor in West Germany as compared with the United States.
37
Household Bargaining and Risk of Divorce
The institutional context shapes more than the household division of labor; it also
varies the effects of different divisions of paid and unpaid labor on marital stability.
Gender specialization proves optimal for marital stability in the country with
institutional support for this model: West German male breadwinner couples are the
most stable in that country and any movement away from this in terms of wives’ relative
earnings or husbands’ relative housework increases the risk of divorce. In contrast,
equitable distributions of the household division of labor predicted under social
exchange models appear optimal in the United States where policy remains silent on the
private sphere and market effects meander their laissez-faire course. At the extremes,
however, both U.S. male and female breadwinner couples are at greater risk of divorce.
Yet female breadwinner couples neutralizing gender deviance by having wives’ perform
an equal share of domestic tasks are more stable than traditional male breadwinner
couples in either country.
These results suggest important extensions of our understanding of gender
relations. First, the competing theories of effects of the household division of labor both
hold true; which holds true depends upon the degree to which theorized dynamics are
supported by institutional factors such as policies. This suggests that the slow evolution
in the division of domestic tasks observed over the past half century may not result from
persistent gender differences, but from continuing institutional reinforcement of the
gendered division of labor. Change is not revolutionary as initially predicted by
Hochschild (1987); policies affect the progress of the evolution toward greater gender
equity.
38
Household Bargaining and Risk of Divorce
Second, where the evolution is not thwarted by policies reinforcing traditional
gender hierarchies, men’s greater participation in domestic tasks results in more stable
marriages regardless of women’s employment. This finding raises questions regarding
the wisdom of calls from conservative quarters to reinforce the traditional male
breadwinner family to turn the tide of increasing family instability. Further, a gendered
division of labor leaves women and their children economically vulnerable under macro
economic conditions that are more volatile than those of the 1960s. For example,
unemployment in West Germany was just 0.5% in 1965, rising during the 1980s and 90s
to 7 or 8% (OECD 1978, 2000). Most vulnerable during periods of high unemployment
are persons with low skills and little experience, such as traditional housewives. Single
female heads-of-household are more likely to be poor and reliant on state transfers (Daly
and Rake 2003). Consequently, reinforcement of the male breadwinner model appears of
little benefit to women, children, or the state.
One might lament that U.S. female breadwinners must carry a one-and-one-half
burden of paid and unpaid labor to ensure marital stability. Women employed full-time,
however, reduce their domestic hours, so such women’s greater domestic share is based
upon fewer total housework hours to be divided between the wife and husband. In
addition, we cannot tell with these data the extent to which performing the remaining
household tasks represents an expression of love and caring rather than a burden (Ferree
1990). Future research needs to decipher, however, what U.S. husbands that contribute
little to either paid or unpaid labor do contribute to marriage to make them more
successful than traditional ones.
39
Household Bargaining and Risk of Divorce
More generally, the results shed no light on the process by which the household
division of labor might lead to marital instability. Does the household division of labor
alter marital quality, or do different divisions only alter the risk of divorce within
unhappy marriages as found by Schoen and his colleagues (2002)? There is also a tacit
assumption within both models that women initiate divorce, either because they no
longer economically require men (specialization and trading model), or because men fail
to take on more equitable divisions of domestic tasks (social exchange model). Given
that different household divisions of labor appear optimal in different countries, our
understanding of gender relations would deepen with exploration into which partner
under what circumstances terminates the relationship under different policy
configurations.
We remain somewhat cautious in our conclusions, however, because even among
the more recent group of first married couples analyzed here, female primary
breadwinning couples are still outliers. Wives earned more than 75% of the income in
only 13% of U.S. and 7% of West German couples in the sample. Still, the results
highlight that distributive justice prevails where gender hierarchies are not reinforced.
Policy is instrumental in setting gender hierarchies, with the evidence here indicating
that policy encouraging gender equity encourages family economic flexibility and
marital stability. In other words, gender equity represents more than a feminist ideal; it
proves essential for sustaining healthy post-industrial societies. Neither the United
States nor Germany, however, has policy provisions actively supporting maternal
employment and encouraging men’s participation in domestic sphere, particularly
40
Household Bargaining and Risk of Divorce
childcare, such as found in Scandinavian countries (Gauthier 2005; Gornick and Meyers
2003). Future comparative research needs to look at the household divisions of labor
and effects of these on family outcomes across a wider array of state, market and gender
relations.
41
Household Bargaining and Risk of Divorce
REFERENCES
Aberg, Yvonne. 2003. “Social Interactions: Studies of Contextual Effects and
Endogenous Processes.” PhD dissertation, Department of Sociology, University of
Stockholm, Sweden.
Allison, Paul D. 1984. Event History Analysis: Regression for Longitudinal Analysis
(Sage University Paper Series in Quantitative Applications in the Social Sciences
No. 07-046). Thousand Oaks, CA: Sage.
Baxter, Janeen. 1997. “Gender Equality and Participation in Housework: A Cross-
National Perspective.” Journal of Comparative Family Studies 28: 220-47.
Becker, Gary S. 1981. A Treatise on the Family. Cambridge, MA: Harvard University
Press.
———. 1985. “Human Capital, Effort, and the Sexual Division of Labor.” Journal of
Labor Economics 3, Part 2: S33-S58.
Becker, Gary S., Elisabeth M. Landes, and Robert T. Michael. 1977. “An Economic
Analysis of Marital Instability.” The Journal of Political Economy 85, 6: 1141-
1188.
Bianchi, Suzanne M., Melissa A. Milkie, Liana C. Sayer, and John P. Robinson. 2000.
“Is Anyone Doing the Housework? Trends in the Gender Division of Household
Labor.” Social Forces 79.1: 191-228.
Bittman, Michael, Paula England, Nancy Folbre, Liana Sayer, and George Matheson.
2003. “When Does Gender Trump Money? Bargaining and Time in Household
Work.” American Journal of Sociology 109(1): 186-214.
42
Household Bargaining and Risk of Divorce
Blair, Sampson L. and Daniel T. Lichter. 1991. “Measuring the Division of Household
Labor: Gender Segregation of Housework Among American Couples.” Journal of
Family Issues 12: 91-113.
Blau, Peter. 1960. Exchange and Power in Social Life. New York: John Wiley and
Sons.
Blau, Francine D., Marianne A. Ferber, and Anne Winkler. 2002. The Economics of
Women, Men, and Work, Fourth Edition. Englewood Cliffs: Prentice-Hall.
Blood, Robert O. and Donald M. Wolfe. 1960. Husbands and Wives. New York: Free
Press.
Brines, Julie. 1994. “Economic Dependency, Gender and the Division of Labor at
Home.” American Journal of Sociology 100: 652-688.
Brines, Julie and Kara Joyner. 1999. “The Ties that Bind: Principles of Cohesion in
Cohabitation and Marriage.” American Sociological Review 64: 333-55.
Budde, Gunilla-Friederike. 1999. “Women’s Policies in the GDR in the 1960s/70s:
Between State Control and Societal Reaction.” Pp. 199-217 in Rolf Torstendahl, ed.
State Policy and Gender System in the Two German States and Sweden 1945-1989.
Lund, Sweden: Bloms i Lund Tryckeri AB.
Bumpass, Larry L., Teresa Castro Martin, and James A. Sweet. 1991. “The Impact of
Family Background and Early Marital Factors on Marital Disruption.” Journal of
Family Issues 12: 22-42.
Cooke, Lynn Prince. 2004. “The Gendered Division of Labor and Family Outcomes in
Germany.” Journal of Marriage and Family 66 (December): 1243-1256.
43
Household Bargaining and Risk of Divorce
Daly, Mary and Katherine Rake. 2003. Gender and the Welfare State: Care, Work and
Welfare in Europe and the USA. Cambridge: Polity Press.
Davis, Shannon N. and Theodore N. Greenstein. 2004. “Cross-National Variations in the
Division of Household Labor.” Journal of Marriage and Family 66: 1260-1271.
Emerson, Richard M. 1962. “Power-Dependence Relations.” American Sociological
Review 27, 1: 31-41.
England, Paula. 1993. “The Separative Self: Androcentric Bias in Neoclassical
Assumptions.” Pp. 37-53 in Beyond Economic Man: Feminist Theory and
Economics, edited by Marianne A. Ferber and Julie A. Nelson. Chicago: University
of Chicago Press.
England, Paula and George Farkas. 1986. Households, Employment, and Gender: A
Social, Economic and Demographic View. New York: Aldine deGruyter.
EUROSTAT. 2000. New Cronos Database 2000.
Fenstermaker Berk, Sarah. 1985. The Gender Factory: The Apportionment of Work in
American Households. New York: Plenum Press.
Ferree, Myra Marx. 1990. “Beyond Separate Spheres: Feminism and Family Research.”
Journal of Marriage and the Family 52(4): 866-84.
Fuwa, Makiko. 2004. “Macro-level Gender Inequality and the Division of Household
Labor in 22 Countries.” American Sociological Review 69: 751-67.
Gauthier, Anne H. 2005. “Trends in Policies for Family-Friendly Societies.” Pp. 95-110
in The New Demographic Regimes: Population Challenges and Policy Responses,
edited by Miroslav Macura, Alphonse L. MacDonald and Werner Haug. New York:
44
Household Bargaining and Risk of Divorce
United Nations.
Gerhard, Ute. 1992. “German Women and the Social Costs of Unification.” German
Politics and Society 24 & 25 (Winter 1991-92): 16-33.
Gershuny, Jonathan. 2000. Changing Times: Work and Leisure in Postindustrial
Society. Oxford: Oxford University Press.
Geschka, Otti. 1989. “Participation and Disadvantage: Women in the Educational
System.” Pp. 189-198 in The Federal Republic of Germany, edited by Eva
Kolinksy. Oxford: Berg Publishers.
Goffman, Erving. 1977. “The Arrangement Between the Sexes.” Theory and Society 4:
301-31.
Goldin, Claudia. 1990. Understanding the Gender Gap: An Economic History of
American Women. New York: Oxford University Press.
Goldscheider, Frances K. and Linda J. Waite. 1991. New Families, No Families? The
Transformation of the American Home. Berkeley, CA: University of California
Press.
Gornick, Janet and Marcia K. Meyers. 2003. Families that Work: Policies for
Reconciling Parenthood and Employment. New York: Russell Sage.
Greenstein, Theodore N. 1990. “Marital Disruption and the Employment of Married
Women.” Journal of Marriage and Family 52: 657-676.
———. 1995. “Gender Ideology, Marital Disruption, and the Employment of Married
Women.” Journal of Marriage and Family 57: 31-42.
———. 2000. “Economic Dependence, Gender, and the Division of Labor in the
45
Household Bargaining and Risk of Divorce
Home: A Replication and Extension.” Journal of Marriage and Family 62: 322-335.
Gupta, Sanjiv. 1999a. “The Effects of Transitions in Marital Status on Men’s
Performance of Housework.” Journal of Marriage and Family 61: 700-711.
Gupta, Sanjiv. 1999b. “Gender Display? A Reassessment of the Relationship Between
Men’s Economic Dependence and Their Housework Hours.” Paper presented at the
1999 annual meeting of the American Sociological Association, Chicago.
Hantrais, Linda. 1994. “Comparing Family Policy in Britain, France and Germany.
Journal of Social Policy 23, 2: 135-160.
Harkness, Susan and Jane Waldfogel. 1999. “The Family Gap in Pay: Evidence from
Seven Industrialised Countries.” Luxembourg Income Study Working Paper #219.
Hartmann, Heidi. 1981. “The Family as the Locus of Gender, Class, and Political
Struggle: The Example of Housework.” Signs 6: 366-394.
Hobson, Barbara. 1990. “No Exit, No Voice: Women’s Economic Dependence and the
Welfare State.” Acta Sociologica 33 (3): 235-250.
Hochschild, Arlie, with Anne Machung. 1989. The Second Shift: Working Parents and
the Revolution at Home. London: Piatkus Ltd.
Hoffman, Saul D. and Greg J. Duncan. 1995. “The Effect of Incomes, Wages, and
AFDC Benefits on Marital Disruptions.” Journal of Human Resources 30: 19-41.
Huber, Joan and Glenna Spitze. 1980. “Considering Divorce: An Expansion of
Becker’s Theory of Marital Instability.” American Journal of Sociology 86(1): 75-
89.
Jaumotte, Florence. 2003. “Female Labour Force Participation: Past Trends and Main
46
Household Bargaining and Risk of Divorce
Determinants in OECD Countries.” Economics Working Paper ECO/WKP 30.
Paris: OECD.
Kamo, Yoshinori. 1991. “A Nonlinear Effect of the Number of Children on the
Division of Household Labor.” Sociological Perspective 34: 205-218.
———. 1994. “Division of Household Work in the United States and Japan.” Journal
of Family Issues 15: 348-378.
Leibfried, Stephan and Ilona Ostner. 1991. “The Particularism of West German Welfare
Capitalism: The Case of Women’s Social Security.” Pp. 164-186 in The Sociology
of Social Security, edited by M. Adler, C. Bell, J. Clasen and A. Sinfield.
Edinburgh: Edinburgh University Press.
Lennon, Mary Clare and Sarah Rosenfeld. 1994. “Relative Fairness and the Division of
Housework: The Importance of Options.” American Journal of Sociology 100: 506-
31.
Lewis, Jane. 1992. “Gender and the Development of Welfare Regimes.” Journal of
European Social Policy 3: 159-173.
Lillard, Lee A. and Constantijn W. A. Panis. 1994. “Panel Attrition for the PSID:
Household Income, Marital Status, and Mortality.” Working Paper Series 94-16,
Rand Labor and Population Program, Santa Monica, CA.
Lundberg, Shelly and Robert A. Pollak. 1993. “Separate Spheres Bargaining and the
Marriage Market.” Journal of Political Economy 101: 988-1010.
Lundberg, Shelly and Robert A. Pollak. 1994. “Noncooperative Bargaining Models of
Marriage.” American Economic Review 84: 132-137.
47
Household Bargaining and Risk of Divorce
Lundberg, Shelly, Robert A. Pollak, and Terence J. Wales. 1997. “Do Husbands and
Wives Pool Their Resources?” Journal of Human Resources 32: 463-480.
Manser, Marilyn and Murray Brown. 1980. “Marriage and Household Decision-
Making: A Bargaining Analysis.” International Economic Review 21: 31-44.
McAllister, Ian. 1990. “Gender and the Division of Labor: Employment and Earnings
Variation in Australia.” Work and Occupations 17: 79-99.
McElroy, Marjorie B. and Mary Jean Horney. 1981. “Nash-Bargained Household
Decisions: Toward a Generalization of the Theory of Demand.” International
Economic Review 22: 333-349.
Moeller, Robert. G. 1993. Protecting Motherhood: Women and the Family in the
Politics of Postwar West Germany. Berkeley: University of California Press.
———. 1997. West Germany Under Construction: Politics, Society, and Culture in the
Adenauer Era. Ann Arbor: University of Michigan Press.
Moffitt, Robert A. 2000. “Female Wages, Male Wages, and the Economic Model of
Marriage.” Pp. 302-319 in The Ties That Bind: Perspectives on Marriage and
Cohabitation, edited by Linda J. Waite. New York: Aldine de Gruyter.
Münz, Rainer and Ralf E. Ulrich. 1995. “Depopulation After Unification? Population
Prospects for East Germany, 1990-2010.” German Politics and Society 13, 4: 1-49.
National Center for Health Statistics. 1995. Monthly Vital Statistics Report 43
(9).Washington, DC: U.S. Department of Health and Human Services.
O’Connor, Julia, Ann Shola Orloff and Sheila Shaver. 1999. States, Markets, Families:
Gender, Liberalism and Social Policy in Australia, Canada, Great Britain and the
48
Household Bargaining and Risk of Divorce
United States. Cambridge: Cambridge University Press.
OECD. 1978. Labour Force Statistics. Paris: Organisation for Economic Co-operation
and Development.
OECD. 2000. Labour Force Statistics. Paris: Organisation for Economic Co-operation
and Development.
Oláh, Livia. 2003. “Gendering Fertility: Second Births in Sweden and Hungary.”
Population Research and Policy Review 22: 171-200.
Oppenheimer, Valerie K. 1988. “A Theory of Marriage Timing.” American Journal of
Sociology 94: 563-591.
———. 1997. “Women’s Employment and the Gain to Marriage: The Specialization
and Trading Model.” Annual Review of Sociology 23: 431-453.
Orloff, Ann Shola. 1996. “Gender in the Welfare State.” Annual Review of Sociology
22: 51-78.
Ostner, Ilona. 1993. “Slow Motion: Women, Work and the Family in Germany.” Pp.
92-115 in Women and Social Policies in Europe: Work, Family and the State, edited
by Jane Lewis. Aldershot: Edward Elgar.
Ostner, Ilona. 1992. “Ideas, Institutions, Traditions: The Experience of West German
Women, 1945-1990.” German Politics and Society 24 & 25 (Winter 1991-92): 87-
99.
Parsons, Talcott. 1942. “Age and Sex in the Social Structure of the United States.”
American Sociological Review 7: 604-616.
49
Household Bargaining and Risk of Divorce
———. 1953. “A Revised Analytical Approach to the Theory of Social Stratification.”
Pp. 92-125 in Class, Status and Power, edited by R. Bendix and S. M. Lipset. New
York: Free Press.
Press, Julie E. and Eleanor Townsley. 1998. “Wives’ and Husbands’ Housework
Reporting: Gender, Class and Social Desirability.” Gender and Society 12(2): 188-
218.
Ruggles, Steven. 1997. “The Rise of Divorce and Separation in the United States,
1880-1990.” Demography 34, 4: 455-466.
Schoen, Robert, Nan Marie Astone, Kendra Rothert, Nicola J. Standish, and Young J.
Kim. 2002. “Women’s Employment, Marital Happiness, and Divorce.” Social
Forces 81, 2: 643-662.
Shelton, Beth A. and Daphne John. 1996. “The Division of Household Labor.” Annual
Review of Sociology 22: 299-322.
Sorensen, Annemette and Sara McLanahan. 1987. “Women’s Married Economic
Dependency, 1940-1980.” American Journal of Sociology 93: 659-687.
South, Scott J. 2001. “Time Dependent Effects of Wives’ Employment on Marital
Dissolution.” American Sociological Review 66: 226-245.
South, Scott J. and Kim M. Lloyd. 1995. “Spousal Alternatives and Marital
Dissolution.” American Sociological Review 60: 21-35.
South, Scott J. and Glenna Spitze. 1994. “Housework in Marital and Nonmarital
Households.” American Sociological Review 59: 327-347.
Sweeney, Megan M. 2002. “Two Decades of Family Change: The Shifting Economic
50
Household Bargaining and Risk of Divorce
Foundations of Marriage.” American Sociological Review 67: 132-147.
Thibaut, John W. and Harold H. Kelley. 1959. The Social Psychology of Groups. New
York: Wiley.
Torr, Berna Miller and Susan E. Short. 2004. “Second Births and the Second Shift: A
Research Note on Gender Equity and Fertility.” Population and Development
Review 30(1): 109-130.
U.S. Bureau of Labor Statistics. 2004. “Employment Characteristics of Families.”
http://stats.bls.gov/news.release/famee.t04.htm.
U.S. Census Bureau. 2000. Statistical Abstract of the United States 2000. Washington,
D.C.
U.S. Department of Education. 2000. Digest of Education Statistics 2000. Washington,
DC: National Center for Education Statistics.
von Oertzen, Christine. 1999. “Women, Work and the State: Lobbying for Part-Time
Work and ‘Practical Equality’ in the West German Civil Service 1958-1969.” Pp.
79-104 in State Policy and Gender System in the Two German States and Sweden
1945-1989, edited by R. Torstendahl. Lund, Sweden: Bloms i Lund Tryckeri AB.
Waite, Linda J. and Lee A. Lillard. 1991. “Children and Marital Disruption.” American
Journal of Sociology 96:930-953.
Weber, Max. 1958. From Max Weber: Essays in Sociology. New York: Oxford
University Press.
West, Candace and Don H. Zimmerman. 1987. “Doing Gender.” Gender and Society 1:
125-151.
51
Household Bargaining and Risk of Divorce
Witte, James C. and Gert Wagner. 1995. “Declining Fertility in East Germany after
Unification: A Demographic Response to Socioeconomic Change.” Population and
Development Review 21(2): 387-97.
Yamaguchi, Kazuo. 1991. Event History Analysis (Applied Social Research Methods
Series Volume 28). London: Sage Publications.
Zimmerman, Klaus. 1993. “Labour Responses to Taxes and Benefits in Germany.” Pp.
192-240 in Welfare and Work Incentives: A Northern European Perspective, edited
by A.B. Atkinson and Gunnar Viby Mogensen. Oxford: Clarendon Press.
52
Household Bargaining and Risk of Divorce
FIGURE 1
PERCENT WEST GERMAN VERSUS U.S. WOMEN AND MEN AGREEING OR
DISAGREEING WITH THE TRADITIONAL GENDERED DIVISION OF LABOR
0 10203040506070809010
Men
Women
Men
Women
0
West German U.S.
Percentage Respondents
Agree Disagree
Note: Based on 1994 International Social Survey Program data. “Agree” indicates the
percentage of respondents that agree or strongly agree with the statement, “A man’s job is to
earn money; a woman’s job is to look after the home and family”; “Disagree” indicates the
percentage of respondents that disagree or strongly disagree with the statement. Percentages do
not add to 100 as percentage of respondents claiming “Don’t know/Can’t choose” are excluded.
53
Household Bargaining and Risk of Divorce
FIGURE 2
WEST GERMAN AND U.S. ATTITUDES TOWARD DIVORCE
1994 ISSP
0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
5
Married people happier Bad marriage better than single Divorce best solution
West Germany United States
Note: Based on 1994 International Social Survey Program data, reflecting mean country scores
on the questions ranging from 1, strongly agree to 5, strongly disagree.
54
Household Bargaining and Risk of Divorce
TABLE 1
MEANS AND STANDARD DEVIATIONS FOR WEST GERMAN AND U.S. COUPLE
INDEPENDENT AND CONTROL VARIABLES USED IN EVENT HISTORY MODELS OF
DIVORCE
WEST GERMANY UNITED STATES
Wife’s share of labor earnings (0-100) . . . . 18.24 32.18
( 25.71) ( 24.12)
Husband’s share of housework (0-100) . . . 28.47 33.05
( 18.82) ( 24.58)
(Husband’s share of housework)2 . . . . . . . . 1165 1697
( 1340) ( 2349)
Wife’s weekly work hours . . . . . . . . . . . . . 13.63 28.15
( 18.15) ( 15.74)
Wife’s weekly hours of housework . . . . . . 28.80 17.40
(15.05) ( 14.78)
Total household income (000). . . . . . . . . . . 74.96 48.85
( 40.66) ( 44.84)
Length of marriage . . . . . . . . . . . . . . . . . . . 5.39 4.57
( 3.43) ( 2.91)
Husband’s weekly work hours . . . . . . . . . . 39.74 41.59
(14.47) ( 13.35)
Wife’s age . . . . . . . . . . . . . . . . . . . . . . . . . 30.25 30.23
( 5.13) ( 5.98)
Number of children . . . . . . . . . . . . . . . . . . 1.65 1.00
( 0.94) ( 1.01)
Wife out of labor force (1 = yes) . . . . . . . . 0.58 0.09
Wife with university (1 = yes) . . . . . . . . . . 0.10 0.25
Husband with university (1 = yes) . . . . . . . 0.19 0.27
Own home (1 = yes) . . . . . . . . . . . . . . . . . . 0.35 0.53
55
Household Bargaining and Risk of Divorce
Wife earns > 50% * Husband
does <= 50% housework (1 = yes). . . . . . . . 0.07 0.08
Equitable division (1 = yes). . . . . . . . . . . . 0.04 0.14
Husband black (1 = yes). . . . . . . . . . . . . . . . n/a 0.24
N (couples) 522 1,490
(couple-years) 4,483 9,633
Note: These statistics are based on the event history file for risk of divorce from year of
marriage, representing valid couple-years in the observation window. Total household income is
in the local currency.
56
Household Bargaining and Risk of Divorce
TABLE 2
COEFFICIENTS FROM DISCRETE-TIME LOGISTIC REGRESSION OF MARITAL DISSOLUTION FROM YEAR OF MARRIAGE
WEST GERMANY UNITED STATES
Model 1 Model 2 Model 1 Model 2
Log-odds RSE Log-odds RSE Log-odds RSE Log-odds RSE
Wife’s proportional earnings (0-100) 0.01* 0.00. 0.01* 0.01 0.02** 0.01 0.02*** 0.01
Wife out of labor force (0=employed) 0.62 0.48 0.60 0.47 0.83* 0.39 0.99* 0.41
Wife’s weekly work hours - 0.00 0.01 - 0.00 0.01 - 0.01 0.01 - 0.00 0.01
Husband’s % housework (0-100) 0.01* 0.00 0.01* 0.01 - 0.04*** 0.01 - 0.04*** 0.01
(Husband’s% housework)2 -- -- 0.0003* 0.00 0.0002 0.00
Wife’s weekly hours housework 0.01 0.01 0.01 0.01 - 0.03* 0.01 - 0.03* 0.01
Number of children - 0.57*** 0.12 - 0.57*** 0.12 0.11 0.13 0.09 0.13
Log of total household income - 0.12 0.09 - 0.12 0.09 - 0.06 0.11 - 0.05 0.12
Wife with college or university - 0.74+ 0.43 - 0.70 0.45 0.13 0.38 0.20 0.39
Husband with college or university - 0.04 0.30 - 0.00 0.30 - 1.38** 0.51 - 1.38** 0.53
(0=less than college or university)
Home ownership (0=rent) - 0.25 0.22 - 0.23 0.23 - 1.23*** 0.31 - 1.23*** 0.31
Wife’s age - 0.01 0.03 - 0.01 0.03 - 0.03 0.03 - 0.03 0.03
Years since marriage 0.03 0.04 0.03 0.04 0.16** 0.05 0.16** 0.05
Black (US) couple (0=White) -- -- -- -- 0.30 0.24 0.31 0.24
Husband does <50% housework*
wife earns>50% - 0.01 0.37 - 2.25** 0.75
Constant - 2.10 1.42 - 2.10 1.42 - 1.92 1.28 - 2.20 1.40
Log-likelihood - 499 - 499 - 378 -369
Wald chi-square 44.44** 44.48*** 107.00*** 114.25***
n (couple-years) 3524 3524 4204 4204
+ p < .10. *p < .05. **p < .01. ***p < .001. (two-tailed tests)
57
Household Bargaining and Risk of Divorce
FIGURE 3
EFFECT OF EARNINGS-HOUSEWORK EXCHANGE ON U.S. AND WEST GERMAN
COUPLES’ LOG-ODDS OF DIVORCE
-1
-0.5
0
0.5
1
1.5
2
2.5
0102030405060708090
Husband's Housework Share & Wife's Relative Earnings
Change in Log-odds of Divorce
100
U.S. Housework & Earnings
Germany Housework & Earnings
Note: Calculated from coefficients in Table 2. The plots reflect the change in the log-odds of
divorce for percentage point changes in U.S. and German husbands’ share of housework
matching wives’ relative earnings, controlling for significant effects of husbands’ housework
share squared, male breadwinner couples, and wives’ mean hours of housework when
employed or out of the labor force.
58
