Center for Population Dynamics
School of Social and Family Dynamics
Arizona State University
Tempe, Arizona, 85287-3701, USA www.asu.edu/cepod



CePoD Working Paper # 08-101



Sexual Frequency and the Stability of Marital
and Cohabiting Unions



Scott T. Yabiku 1
Constance T. Gager 2





January 2008


Copyrights are held by the authors. This paper may not be cited or quoted without permission.


1 Center for Population Dynamics, School of Social and Family Dynamics, Arizona State University.
2 Montclair State University, Department of Family and Child Studies

---

Sexual Frequency and the Stability of Marital and Cohabiting Unions
Abstract
Prior research on marriage shows that lower sexual frequency or lower sexual satisfaction is
associated with higher rates of divorce. Scant research, however, has addressed the role of sexual
activity in the dissolution of cohabiting unions. Researchers have shown that marriage and
cohabitation are different institutional family forms. Thus, there are good reasons to expect that
the link between sexual activity and stability will differ across marriage and cohabitation. We
draw upon social exchange theory to develop our hypotheses. Our theoretical framework
proposes several reasons why sexual frequency is more important in cohabitation: (1) cohabitors’
lower costs of finding sexual alternatives, (2) cohabitors’ lower barriers of ending the
relationship in the form of union-specific economic and non-economic capital, and (3)
cohabitors’ higher demands for sexual activity. In other words, sexuality occupies a more
prominent role in cohabitation than marriage, and low sexual frequency within cohabitation is
more likely to lead to dissolution. Using the National Survey of Families and Households
(NSFH) and discrete-time event history models, we examine the relationships between sexual
frequency and union dissolution. Results indicate that low sexual frequency is associated with
significantly higher rates of union dissolution among cohabitors than married couples.

---

Sexual Frequency and the Stability of Marital and Cohabiting Unions
Introduction

Prior research finds that sexuality within marriage is an important component of marital
quality and stability. Typically, studies find that higher sexual satisfaction or frequency is
positively associated with marital stability (Yeh, Lorenz, Wickrama, Conger, & Elder 2006;
Edwards & Booth, 1994, Oggins, Leber, & Veroff, 1993; Veroff, Douvan, & Hatchett, 1995;
White and Keith, 1990). The relationship between sexual activity and union stability in
cohabitations has received less research attention, but the existing research finds that sexual
satisfaction in nonmarital unions also promotes stability (Sprecher, 2002).

Although it has previously been established in separate studies that sexual activity is
associated with union stability for both marriage and cohabitation, no studies have compared the
importance of sexual activity across marital and nonmarital unions. Because partners in these
two types of unions have different expectations, histories, and responsibilities (Giddens, 1992),
there is good reason to believe that sexuality within these unions may have differential
stabilizing roles. The role of sexual frequency in relationship satisfaction is an important area of
study as recent research finds that sexual relations ranked as the second most problematic issue
(after balancing job and family) among a national sample of young married couples (Risch,
Riley, & Lawler, 2003).

In this paper, we develop and present a theoretical framework that links sexual activity to
union stability. We examine how the importance of sexual activity varies across marital and
cohabiting unions. We use the first and second waves of the National Survey of Families and
Households (NSFH) to empirically test our hypotheses.
1

---

Theoretical Issues

It has been well-established in the literature that positive, healthy sexuality within
marriage is associated with several dimensions of marital well-being including marital
satisfaction and happiness (Edwards & Booth, 1994; Henderson-King & Veroff, 1994; Perlman
& Abrahmson, 1982; Blumstein & Schwartz, 1983). Research has also found that low sexual
satisfaction can promote marital instability (Edwards & Booth, 1994; Oggins, Leber, & Veroff,
1993; Veroff, Douvan, & Hatchett, 1995; White & Keith, 1990). For example, Edwards and
Booth (1994) found that declines in sexual satisfaction among married couples from 1980 to
1983 were associated with a higher likelihood of divorce 5 years later. White and Booth (1991)
wrote that reports of sexual problems among married couples increased the likelihood of divorce,
net of other relationship quality variables. Fewer studies have focused on the role of sexual
satisfaction in cohabiting or dating relationships. One exception is Sprecher’s (2002) study of the
quality and stability of dating couples. She finds that couples who reported higher sexual
satisfaction scores were more likely to stay together than couples with lower scores.
Social and Biological Factors Linking Sexual Frequency and Dissolution

There are several explanations linking lower sexual frequency or lower sexual
satisfaction to higher rates of union dissolution. One explanation is selection. It may not be that
reduced sexual activity causes union dissolution, but that as partners experience other non-sexual
problems and difficulties in the relationship, their level of intimacy and sexual activity drops as
well. Findings from prior studies are consistent with this reasoning. For example, alcoholism is
likely to cause both low sexual frequency and higher rates of dissolution (O’Farrell, Choquette,
& Birchler, 1991). Also, much research has demonstrated that poor communication is linked to
low levels of sexual satisfaction. Poor communication may also account for higher rates of
2

---

relationship dissolution (Thachil & Bhugra, 2006). In sum, this literature suggests that any
research studying the relationship between sexual activity and union stability must control for
potential confounding causes of both sexual frequency and dissolution in order to avoid spurious
associations.
A second explanation of the relationship between sexual frequency and union stability is
a causal one: the sexual act promotes social attachment between participants. Research from both
the biological and social sciences also is consistent with this explanation. The social sciences
usually focus on behavioral models of attachment, while biological sciences have investigated
neuroendocrine models (Carter, 1998). Neuroendocrine models of attachment point to
neurochemical mechanisms that aid the formation of social bonds. Neurochemicals such as
Oxytocin and vasopressin, for example, are released during sexual activity and may help to
increase social attachment (Carter, 1998; Insel, 1997). While these processes have not been
conclusively identified in humans, animal studies strongly suggest that neurochemical released
during sex do cause social attachment to increase through biochemical mechanisms in the brain
(Insel, 1997).
From a social science perspective, the relationship between sexual activity and union
dissolution can be studied with social exchange theory. Social exchange theory has been used to
analyze a broad range of social interactions (e.g., Blau, 1964; Homans, 1961; Sprecher, 1998),
based on the assumption that in a given interaction, each individual gives something and gets
something in return. Analyzing the costs and benefits of various interpersonal behaviors provides
a useful basis for making predictions about how people will choose to act. Specifically, social
exchange theory can help explain how sexual interaction occurs or does not (Lawrance & Byers,
3

---

1995), and more generally how sexual interactions may influence relationship disruption
decisions (Levinger & Moles, 1979).
Capitalizing on the strengths of exchange theory, Lawrance and Byers (1995) developed
the Interpersonal Exchange Model of Sexual Satisfaction (IEMSS). This model takes into
account how potential partners weigh the rewards and costs of their sexual relationship. Rewards
emerge when exchanges are pleasurable while costs are generated, for example, when exchanges
require physical or mental effort or those that produce pain, embarrassment, or anxiety (Thibaut
& Kelley, 1959). The IEMSS also accounts for an individual’s comparison level (CL) -- the
standard against which individuals judge the attractiveness of their rewards and costs (Kelley &
Thibaut, 1978). In other words, the individual evaluates the level of rewards and costs that they
expect to receive in a sexual relationship. In sum, social exchange theory holds that as the costs
of a given interaction, for example having sex, begin to outweigh the rewards, an individual will
not engage in the interaction.
More generally, social exchange theory can be used to study the relationship between low
sexual frequency and union dissolution. Social exchange theory suggests that partners weigh the
costs and benefits of a marriage in deciding whether to dissolve it or not. Part of this calculation
includes the barriers to ending their current union (Levinger & Moles, 1979). When the costs of
staying in a relationship consistently outweigh the rewards, and barriers are seen as
surmountable, individuals may seek to end the relationship. When faced with low sexual
frequency, we argue that cohabiting couples face fewer barriers to and lower costs of leaving a
relationship. Using social exchange theory, we generate several hypotheses to predict why the
impact of sexual frequency on union stability will differ for cohabitation and marriage.
Differences between Marital and Cohabiting Unions
4

---

To better understand differences between cohabiting and marital unions in relation to
sexual frequency and dissolution, we first need to explicate the broader differences between
marriage and cohabitation. Broadly, marriage and cohabitation involve different levels of
institutionalization. Specifically, cohabitation has been described as an incomplete institution
compared to marriage (Brown & Bulanda, 2005; Smock & Gupta, 2002). The concept of an
incomplete institution was first used by Cherlin (1978) to argue that remarried families with
children lack a clear set of norms and thus proscriptive solutions to problems that emerge. For
example, the role of a step-parent in rearing step-children is less clear than the role of a
biological parent. Nock (1995) extended this idea to argue that cohabitation also falls within the
definition of an incomplete institution. The norms about marriage are clearer and more specific
than those surrounding cohabitation, because cohabitation is a much newer relationship form,
and thus not governed by a clear set of consensual norms. Because the rules and norms
governing cohabiting relationships are less clear, it has been suggested that partners face greater
negotiations over their roles in the relationship than married couples (Brown & Bulanda, 2005),
although direct evidence of this is scant.
Perhaps even more important Nock (1995) noted that cohabitation is not governed by
formal law as marriage is. This partially explains why marriage and cohabitation have different
time horizons (Waite & Joyner, 1992). While both marriage and cohabitation are viewed as
monogamous relationships, cohabiting involves less long-term commitment compared to
marriage as cohabitations are much more easily dissolved. Married couples expect to and do stay
together longer than cohabiting couples (Bumpass & Lu, 2000; Bumpass, Sweet, & Cherlin,
1991). Thus, cohabitation, unlike marriage, carries no explicit social or legal commitment to stay
together for the long term.
5

---

Given these broad differences between marriage and cohabitation, we propose several
hypotheses to explain why the impact of sexual frequency on union stability differs between
marriages and cohabitations. First, cohabiting couples have lower costs of ending the union with
regards to sexual alternatives. When a marriage or cohabitation ends, sexual activity with the
partner also ends: one of the costs of dissolving the union is lost sexual activity. Partners may
seek to replace this lost sexual activity with a new partner, but there are search costs to finding a
new partner (Oppenheimer, 1988). These search costs are probably lower for cohabiting than
married individuals (Kravdal, 1999). Dolcini et al. (1993) reported that while only 2% of married
individuals had more than one sexual partner in the past year, 12% of cohabitors did. Other
research showed that cohabitors are twice as likely to have been unfaithful in the past year (Treas
& Giesen, 2000). We hypothesize that cohabitors will have more ready access to a replacement
sexual partner, which lowers the costs of ending the union with regards to lost sexual activity.
Second, cohabitors have fewer barriers to ending the union because they have fewer
shared investments in economic and non-economic union-specific capital. Children can be
viewed as a form of non-economic, union-specific capital. Cohabiting couples are less likely to
have children present in the household compared with married couples. For example, 39 percent
of cohabiting couples have children compared to 45 percent of married couples who have
children present (U.S. Bureau of the Census, 2003). However, in married couples with children,
the children are more likely to be the biological offspring of both parents compared with children
in cohabiting families. In sum, it is still the case that cohabitors are less likely to have children
than married couples, and that a larger proportion of these cohabiting couples did not have these
children together. Thus, compared to married couples, cohabitors have fewer barriers to dissolve
the union because of a higher likelihood that no children are present or that only one of the
6

---

partners is the child’s biological parent. In contrast to non-economic union-specific capital, home
ownership represents economic union-specific capital. A home represents a substantial economic
asset that would have to be divided if the union were to end (Rindfuss & Van den Heuvel, 1990).
In general, cohabitors do not pool together financial resources to the same degree as married
partners (Morrison & Ritualo, 2000). Thus, married couples are more likely to face substantial
barriers such as selling a large asset or losing touch with biological children than are cohabitors.
In terms of social exchange, while low sexual frequency may lower the benefits of the union for
both marriage and cohabitation, married individuals will have higher barriers to leaving the
relationship, and the impact of low sexual frequency is weakened.
Third, cohabitors might have higher expectations and demands for sexual activity than
married partners. The literature suggests multiple dimensions on which cohabitors’ values,
expectations, and norms differ from married spouses. For example, cohabitors are usually more
individualistic than people in marital unions (Teachman, 2003), and these individualistic
tendencies can interfere with the development of commitment to the relationship and its intrinsic
rewards (Scanzoni, Polonko, Teachman, & Thompson1989). An additional dimension on which
cohabitors and married partners might differ is the expectation of sexual frequency. Researchers
find that cohabitors report higher sexual frequency per month compared to married couples
(Michael, Gagnon, Laumann, & Kolata, 1994). For example, while over 40 percent of married
couples report having sex 2 - 3 times per week, well over 50 percent of cohabiting couples have
sex 2 - 3 times per week (Michael, et al. 1994). Prior research on marital sexual frequency
documents that sexual activity declines with age and relationship duration (James, 1974; Jasso,
1985; Klusmann, 2002; Rao & DeMaris, 1995; Udry 1993; Udry & Morris, 1978; Westoff,
1974). For example, Klusmann (2002) reports that sexual activity and sexual satisfaction decline
7

---

among women and men as the duration of partnership increases and that sexual desire declines in
women as well. Overall, norms and expectations for sexual activity are lower for marriages than
cohabitations. We expect that the disruptive impact of low sexual frequency in marriage will be
weaker than in cohabitation, where partners expect higher levels of sexual activity.
In sum, we expect that higher sexual frequency will be associated with lower rates of
union disruption in both marriage and cohabitation. We hypothesize, however, that higher sexual
frequency will have a stronger effect at maintaining union stability in cohabitation than marriage.
Potential reasons include cohabitors’ lower costs of finding sexual alternatives, cohabitors’ lower
barriers of ending the relationship in the form of union-specific economic and non-economic
capital, and cohabitors’ higher demands for sexual activity.
Data and Methods

To test our hypotheses, we use data from the first two waves of the National Survey of
Families and Households (NSFH). Wave 1 of the NSFH collected a variety of family, household,
and demographic data from a nationally representative sample of individuals in 1987-1988.
NSFH data were collected from a randomly selected adult in each household surveyed and from
the respondent’s spouse or partner. Respondent data were collected through both face-to-face
interviews and self-administered questionnaires; spouses and partners were asked to complete a
shorter, less detailed questionnaire (Sweet, Bumpass, & Call, 1988). Wave 2 data were collected
in 1992-1994, and wave 3 data were collected in 2001-2003, but did not follow up with all
respondents. Thus, our analysis is based on the wave 1 sample of married and cohabiting
respondents who were reinterviewed at wave 2. This includes 5440 marital unions and 328
8

---

Document Outline

• CePoD_WP_cover.pdf
• yabiku_2008.pdf

---