In contrast to the intense public discussion and concern regarding adolescents’ sexual behavior and reproductive health needs, limited attention has been given to the sexual behavior of single adult women. However, demographic shifts in the United States make the behavior and needs of this group increasingly salient. The median age at first marriage rose from 22.0 years in 1980 to 25.3 years in 2002;1 as of 2002, 40% of women aged 25–29 had never married.2 Cohabitation is becoming a common alternative to marriage,3 and the relative impermanency of both marriage and cohabitation means that a woman may be single at different periods in her lifetime. Sexual activity among single women puts them at risk of unplanned pregnancy, unplanned births and STDs, including HIV, and determines the extent of their need for sexual and reproductive health information and services. Given the less stable nature of single women’s sexual relationships, their needs for services may be greater than those of married or cohabiting women.
A better understanding of single women’s sexual and reproductive lives is required to inform public policies that address their behaviors and needs. “Abstinence until marriage” has been a foundation of federal policy for the past decade, but more recently, an unprecedented emphasis on promoting abstinence among people in their 20s has developed.4 Guidelines for the $50 million federal abstinence education grant program to states have expanded the target population to include unmarried adults up to 29 years old.5 Programs utilizing these funds are required to teach that sex among unmarried adults is not only nonnormative, but also unhealthy and destructive, and likely to cause harmful physical and psychological effects.
Additional emphasis on abstinence among adult women is evident in the program priorities for Title X, the only federal program that provides designated funding for family planning. Title X plays an important role because it establishes standards in publicly funded family planning service provision. In July 2003, the Office of Population Affairs announced several new goals for the program, including providing “extramarital abstinence education and counseling” designed to “encourage abstinence outside a mutually monogamous marriage or union.”6(p.41116) It also called upon Title X–funded clinics to “incorporate the ‘ABC’ message” in the integration of family planning and HIV prevention services: “For adolescents and unmarried individuals, the message is ‘A’ for abstinence” (emphasis added).6(p.41117) Family planning advocates have asked how Title X–supported providers are expected to reconcile this requirement with the program’s historical and ongoing mandate to provide contraceptive methods and services.7 Policies that seek to promote abstinence among unmarried individuals need to be informed by the reality of current patterns of sexual behavior and union formation in the United States.
Given demographic trends and the growing policy interest in single women, it is important to understand the sexual behavior of single adult American women, including how it compares with that of married and cohabiting women. While marital and union status is often used as a differentiating variable in studies of sexual behavior, it generally is not the focus of these studies; as a result, information about the sexual behavior and reproductive health needs of single women is scattered across diverse studies.6,8,9 The generalizability of this past work and its contribution to informing current policy are limited by a number of factors. First, earlier studies used various and noncomparable definitions of union status, particularly in how they identified and grouped women in cohabiting unions. Measures that group all unmarried women—whether cohabiting or single—do not accurately capture their experiences. Second, most studies that focused on union status as a differentiating characteristic gave little attention to identifying variation in sexual activity and reproductive health needs by social and demographic characteristics, such as age, income, education, and race or ethnicity. Third, research based on data from the mid-1990s or earlier may no longer adequately describe the experiences and needs of single women.
This article presents new, nationally representative data on current patterns of sexual behavior, by union status, among women aged 20–44. Our objectives are to assess the extent to which single women are sexually active and at risk of poor sexual and reproductive health outcomes, and therefore are in need of reproductive health services; to identify differences in these patterns between single women and married or cohabiting women; and to examine differences in these patterns among demographic subgroups.
We address a range of questions: What proportions of single women have had sexual intercourse, are currently sexually active and have had multiple partners in the past year? What proportions are at risk of unintended pregnancy and STDs? What are their needs for sexual and reproductive health services, and do they have health insurance to help meet these needs? How do single women differ from cohabiting or married women regarding sexual behavior and need for information and services? Are some subgroups of single women at greater reproductive health risk than others?
Data and methods
Most of the data for this analysis were drawn from the 2002 National Survey of Family Growth (NSFG), the latest cycle of a periodic survey of the noninstitutionalized population in the United States. The survey used a multistage, stratified, clustered sample design and interviewed men and women of reproductive age; methodological details are available elsewhere.10 The NSFG used face-to-face interviews to collect information about sexual behavior and family formation, including current union status. A month-to-month calendar was employed to elicit detailed responses about sexual activity and contraceptive use. Because of our focus on adult women, our sample was limited to 6,493 female respondents aged 20–44 at the time of interview. To examine trends in union status, we also used 1988 and 1995 NSFG data on 7,216 and 9,451 women, respectively, who were aged 20–44 at the time of interview. These earlier surveys had comparable designs to the 2002 NSFG.11,12
•Union status. Women were asked to identify their marital status from the following choices: “married; not married but living together with a partner of the opposite sex; widowed; divorced; separated, because you and your husband are not getting along; never been married.” We categorized women who gave the first two responses as married and cohabiting, respectively, and the rest as single. We recognize that this last category encompasses groups that may differ in behaviors and needs, but for our assessment of how women not in a union differ from others, this categorization is appropriate.
•Sexual behavior. We examined a number of women’s sexual behaviors as a way to measure risk of unintended pregnancy or STDs. Most of the measures for this analysis were drawn from the face-to-face interviews, and sensitive topics, such as number of partners, thus may be underreported.13 However, we do not know the extent to which any possible underreporting varies by union type.
Sexual experience is a dichotomous measure indicating whether a woman had ever had vaginal intercourse.* Being sexually active is a dichotomous measure of whether a woman had had vaginal intercourse in the past three months; women were also asked the number of months in which they had had intercourse in the past year (range, 0–12).†
Several variables measured exposure to reproductive health risks and use of contraceptives. Having multiple sexual partners—a risk factor for STDs—was defined as having had two or more partners in the past year. A woman was considered to be at risk of unintended pregnancy if she was sexually active, fecund, not pregnant or postpartum, not trying to get pregnant and not using contraceptive sterilization; women using other forms of contraception were considered to be at risk of unintended pregnancy. Data for both of these measures were collected for all women, regardless of sexual experience, since never having sex is one means of risk reduction. Among women with multiple partners in the past year, we assessed if they had used condoms during the month of interview, as well as their consistency of condom use in the past year (always, sometimes, never). For women who were at risk of unintended pregnancy, we determined if they had used any contraceptive in the month of interview.
•Access to care. We measured one important aspect of women’s difficulty in meeting their reproductive health care needs: not having health insurance. Respondents were asked whether they had had private insurance, Medicaid coverage or other types of insurance during the entire 12 months preceding the survey. Unfortunately, the NSFG has no other appropriate measure of access or barriers to health care.
•Demographic characteristics. We examined several key demographic characteristics: age at interview (in five-year groups), race or ethnicity (white, black or Hispanic),‡ and household poverty status, as a percentage of the federal poverty level (less than 100%, 100–199%, 200% or higher).§ Additionally, we categorized women by the highest level of education completed (less than high school, high school or college); we excluded women younger than 25 to avoid any bias from women who were still in school.
We first examined change in the distribution of women by current union status, according to age, across the 1988, 1995 and 2002 surveys. In the rest of the analyses, we used only 2002 data. We assessed differences in distribution by union status according to age and race or ethnicity, as well as differences in sexual behavior and risk of unintended pregnancy or STDs among single, married and cohabiting women. Among single women, we examined sexual behavior and risk by age, race or ethnicity, poverty status and education level. We also analyzed union status and contraceptive use among women who were at risk of unintended pregnancy or STDs. Finally, we looked at differences in insurance coverage among sexually active women.**
In all analyses, we use paired t tests to assess differences between groups at p≤.05. Standard errors and significance were calculated using the svy series of commands in Stata 8.2 to account for the stratified survey design.14
Characteristics and sexual behavior of single women
A large proportion of American women aged 20–44 are currently single—36%, or 18.6 million women in 2002—and this proportion has not changed since 1988. Rates of sexual experience among single women have also been stable: 86% in 1988, and 89% in 1995 and 2002.
The proportion of women who are single differs significantly by race or ethnicity (Table 1, page 28). The majority of 20–44-year-old black women are single (58%), and only at ages 40–44 does this proportion fall below half (48%). The proportion who are single is similar among all white and Hispanic women (32%). A higher proportion of whites aged 20–24 than of Hispanics in that age-group are single (62% vs. 47%). Nevertheless, similar proportions of Hispanic and white women in each of the older age-groups are single; these proportions stabilize at about one in four among women aged 30 or older.
Compared with both married and cohabiting women, single women are significantly less likely to be sexually experienced (Table 2, page 29). And among sexually experienced women, single women are the least likely to be currently sexually active, and are sexually active for the fewest months in the year. Overall, 89% of single women are sexually experienced, and 70% of this subgroup are sexually active, compared with 95–96% of married or cohabiting women. On average, sexually experienced single women had intercourse in seven of the last 12 months, compared with 11 months for married or cohabiting women. These patterns of differences by union status hold across demographic characteristics.
Among single women, sexual behavior varies significantly by age, race or ethnicity, and education level. Single women in their 20s are less likely than older women to be sexually experienced (83% vs. 94–95%). But among those who have ever had sex, the proportion who are sexually active declines with age, from 78% of women in their 20s to 56% of those aged 40–44. The average number of months that single women have been sexually active in the past year also declines significantly with age, from eight months among those in their 20s to six months among those aged 40–44.
Black single women are more likely than whites to be sexually experienced (94% vs. 88%) and sexually active (74% vs. 68%), and the average number of months they have been sexually active in the past year is greater (eight vs. seven months). Among women aged 25 or older, lower proportions of single women with a college education than of those with a high school education are sexually experienced (86% vs. 96%) and sexually active (55% vs. 67%). College graduates also report fewer months of sexual activity over the last year than do high school graduates (six vs. seven months), while women with less than a high school degree report the most months of sexual activity (eight). Household poverty status is not significantly related to any of these measures.
STD and unintended pregnancy risk
•Multiple sexual partners. Single women are more likely to have had two or more sexual partners in the past year (22%) than are either cohabiting or married women (9% and 2%, respectively—Table 3). The same is true in every demographic subgroup examined.
The proportion of single women reporting multiple partners declines with age, from 26% of 20–29-year-olds to 15% of 40–44-year-olds. The likelihood of having had multiple partnerships is greater among white than among Hispanic single women (23% vs. 18%), and greater among women with a high school degree than among those having a college degree (22% vs. 16%). Multiple partnership does not differ by household poverty status (21–23%).
•Risk of unintended pregnancy. Overall, a smaller proportion of single women than of cohabiting women are at risk of unintended pregnancy (49% vs. 60%). Risk does not differ significantly between single and married women, but does differ between cohabiting and married women (p≤.05—not shown).
Within most subgroups, single women are less likely than cohabiting women to be at risk of unintended pregnancy. The proportions of single and cohabiting women at risk differ significantly at ages 20–39, but not at age 40–44. Among blacks and whites, but not among Hispanics, single women are at greater risk of unintended pregnancy than are married women.
The proportion of single women at risk of unintended pregnancy declines with age, from 63% to 25%. It does not vary significantly by race or ethnicity, but is lowest among those with the least education. Of the measures we examine, unintended pregnancy risk is the only to vary by poverty status among single women: Those who have household incomes of 200% or more of the federal poverty level have a higher rate of risk than their lower income counterparts (55% vs. 42–45%).
•Condom and contraceptive use. Among women who have multiple partners, consistent condom use is important for protection against STDs. However, only 28% of these women had used condoms in the past month, and higher proportions of single than of married women had recently used condoms (30% vs. 11%) and had used them consistently over the last year (19% vs. 3%—Table 4). Yet more than a quarter of single women with multiple partners had never used condoms over this period, and more than half had used them inconsistently.
Among all women at risk of unintended pregnancy, higher proportions of married or cohabiting women than of single women had used a contraceptive in the past month (86% vs. 81%). Overall, one in six women at risk had not used any method.
•Union status and risk. We also assessed whether single women were disproportionately at risk of unintended pregnancy or STDs. Single women—who make up 36% of women aged 20–44—represent 27% of those who are sexually active, 82% of those with multiple partners (of whom they are a disproportionately large part) and 37% of those at risk of unintended pregnancy (Figure 1). In contrast, married women make up 54% of 20–44-year-olds and represent 62% of sexually active women, 9% of women with multiple partners and 52% of those at risk of unintended pregnancy. Only 10% of women in this age-group are cohabiting, and they account for 9–12% of each risk subgroup.
•Insurance coverage and risk. In general, health insurance—whether private or Medicaid—provides coverage for needed reproductive health care services, including family planning, maternity care, and STD prevention and treatment services. Most sexually active women, regardless of union status, have health insurance; only 16% did not have insurance throughout the 12 months preceding the survey. However, among sexually active women, the level of uninsurance is significantly higher among cohabiting (25%) and single women (21%) than among married women (12%); the difference between cohabiting and single women is not significant. Cohabiting and single women also are significantly more likely than their married peers to have Medicaid coverage and less likely to have private insurance. These findings suggest that sexually active single women are less able than married women to obtain the sexual and reproductive health services they may need.
Much of the academic and policy discussion of changes in American fertility and family formation has focused on the shift away from formal marriage toward higher rates of cohabitation.3,15–18 Yet one-third of adult women of reproductive age are not currently married or cohabiting, and for the majority of adult women, living without a partner does not mean living without sex. Nine in 10 single women are sexually experienced, and seven in 10 of these experienced women have had intercourse in the past three months. These high levels of sexual activity among single women highlight their need for reproductive health services and their potential risk for poor health outcomes.
How can single adult women achieve healthy sexual relationships while limiting their risk of unintended pregnancy and STDs? Women who are not in a union merit special attention because, as indicated by our results, their behaviors differ significantly from those of married or cohabiting women. It is particularly challenging for women in shorter term or sporadic relationships to maintain effective contraceptive use during all periods when they are at risk of unintended pregnancy, even when they wish to prevent pregnancy.17,18 Providers should be aware of the relationship patterns of single women and offer appropriate counseling on method choices that addresses women’s needs for follow-up and continuity of care.
Single women are more likely than married or cohabiting women to have multiple sexual partners, and this differential pattern by union status corroborates previous findings based on older data.19 The higher level of multiple partnerships among single women puts them and their partners at elevated risk of contracting STDs. Moreover, four of five women with multiple partners are single, so the need for STD prevention services is concentrated among these women. Hence, it is critical to educate single women about their risk for STDs, to increase their use of barrier methods, and to improve their communication and negotiation skills with new partners. Additionally, compared with younger women who are single, older, formerly married women may have less experience with barrier methods and may need greater education and counseling about the importance and use of condoms.20
Among single women, sexual experience, sexual activity, multiple partnerships and number of months sexually active all vary by level of education. In contrast, household income shows little correlation with their sexual behavior. This is somewhat surprising, given the body of research suggesting that income is related to union formation.21 Perhaps this association with union formation and not sexual behaviors indicates that income and poverty status are more influential in longer term than in shorter term decision making.
About half of sexually experienced single women aged 40–44 are sexually active, and on average, these women have sex in only six months of the year. Do these patterns represent differences in personal preferences between older and younger women, or barriers to finding a partner? Challenges to forming intimate long-term partnerships may have implications for older women’s health and well-being.22
Our findings suggest that other groups, beside single women, also merit special attention. One key finding is that a higher proportion of cohabiting women than of either married or single women are at risk of unintended pregnancy. Indeed, the proportion does not differ between married and single women. Cohabiting women are as likely to be sexually active as married women, but less likely to have health insurance. Publicly funded family planning services could be an important resource for cohabiting women; among women who received at least one reproductive health service from a family planning clinic in 2002, 34% were married, 16% were cohabiting and 50% were unmarried.8(Table 18) A 2004 review of Title X–funded family planning services supports the development and testing of approaches for serving couples with the goal of helping them become effective contraceptive users.23 Providers may want to expand their efforts to reach out to cohabiting couples and meet their unique reproductive health and contraceptive needs.
Another high-need group identified in our analysis are single women aged 20–29. The needs of this relatively young group are often overlooked. In some respects, this group is at greater risk than teenage females, because a higher proportion are sexually experienced, they are sexually active for more months in the year and they are less likely to receive health insurance coverage through their parents. Compared with older women, 20–29-year-olds are more likely to be sexually active, have multiple partners and be at risk of unintended pregnancy. New policy and funding emphases on promoting abstinence among single women in their 20s brings attention to this group, but does little to meet their existing reproductive health needs.
Proposed policies that promote abstinence until marriage are of special concern for black women, whose rates of marriage are particularly low; in 2002, only three in 10 black women aged 20–44 were married, compared with more than half of white or Hispanic women. Federal, state and private initiatives to promote marriage seek to address a range of issues perceived as barriers to marriage, especially among low-income minority populations.24-26 Congress has budgeted up to $200 million per year for fiscal years 2006–2010 for marriage promotion programs, in addition to a similar level of federal and matching state funding for abstinence-only programs.27
This study has a number of limitations. First, self-reported data may be inaccurate or incomplete, especially for sensitive topics such as number of sexual partners and sexual activity. Second, among single women, there is likely a range of unidentified relationship types that may be relevant for determining their risk of STDs or unintended pregnancy, as well as their need for reproductive health services. Finally, the data are cross-sectional and provide information about a limited time period prior to the survey. Further research should consider how past unions and changes in union status influence current sexual behaviors and contraceptive use.
The call for a national cultural transformation to sex only within marriage faces the challenge of altering the behavior of the majority of Americans. Since the 1950s, about 90% of Americans have engaged in premarital sexual activity.28 And with the median age at first marriage now around 25, and with single adult women reporting high levels of sexual activity, the behavioral changes required for current efforts to promote abstinence until marriage to succeed would have to be phenomenal. At best, federal and state policies that fail to take into account the reality of women’s sexual behaviors and health needs are unlikely to be effective, and at worst they may have a negative impact on women’s sexual and reproductive health.