Search tips
Search criteria 


Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
Sex Roles. Author manuscript; available in PMC 2010 May 17.
Published in final edited form as:
Sex Roles. 2009 October; 61(7-8): 536–553.
PMCID: PMC2871695

When, where, why and with whom homeless women engage in risky sexual behaviors: A framework for understanding complex and varied decision-making processes


Impoverished women worldwide are at high risk for contracting HIV/AIDS. This study explores how homeless women make risky sex decisions and the role that alcohol and drugs play in this process. We analyze 56 in-depth qualitative descriptions of recent sexual episodes among 28 women living in shelters in Los Angeles, California, USA. The sample (age 18–63) was 46% African American, 21% Hispanic/Latina, and 21% white. Findings suggest that: (1) homeless women engage in multiple types of relationships and sexual behaviors; (2) emotion and attachment play critical roles in women’s risky sex choices; and (3) the role of alcohol and drugs on such choices varies across relationship commitment. Understanding the complexity of sexual decision-making among this population has implications for developing successful risk reduction interventions.

Keywords: Homeless women, risky sexual behavior, decision-making, alcohol and drug use, condom use


Throughout the world, it is generally acknowledged that impoverished women are at higher risk for contracting HIV/AIDS than their more affluent counterparts. To better understand how such women make decisions about risky sex during sexual encounters, we conducted in-depth qualitative interviews with 28 homeless women in the U.S. about their recent sexual encounters. This exploratory study is based on women’s qualitative descriptions of 56 sexual events. In each, women report in their own words what they did, how they felt, and what they were thinking immediately before, during, and after sexual intercourse. Each account provides details on the kinds of events that led up to the sexual activity, the specific sexual activities women engaged in, their condom use or lack of use, the relationships they had with their partners, the physical and social context in which the sexual encounter occurred, and how they came to make decisions about engaging in sex and using (or not using) condoms, including the role that drugs and alcohol might have played in the process. By examining sexual behavior at the level of specific sexual events, this study provides insights into the complexity of sexual decision-making for women of limited resources.

Homeless women are some of the most vulnerable and poverty-stricken persons in the nation. Studies have found that these impoverished and disenfranchised women are more likely to experience a number of threats to their health and well-being. Homeless women are more likely to use alcohol and drugs, experience victimization by violence and, in part due to their need to survive difficult circumstances, engage in unprotected sex and other risky sexual behaviors that may lead to infection by HIV and other STDs (Bassuk et al., 1996; Wenzel et al., 2007; Wenzel et al., 2004; Wood et al., 1990). For homeless women, protection against HIV infection may be a particular challenge given their generally harsher living conditions and reduced access to health care compared to housed and higher income women (Milburn & D'Ercole, 1991; Nyamathi et al., 2000; Padgett & Struening, 1992). Condoms may not always be available, or condoms may be less likely to be used when a woman is more heavily invested in the relationship (Tucker et al., 2007) or when the partner is abusive (Pulerwitz, 2000; Wingood et al., 2000). Evidence is mixed as to whether women are less likely to use condoms when one or both partners are influenced by substance use (Leigh, 2002). For this paper, we examine this and other issues for homeless women.

In this exploratory study, we address two fundamental questions: (1) How do women staying in temporary shelter settings (herein referred to as “homeless”) make decisions about engaging in risky and non-risky sex?; and (2) What role do drugs and alcohol play in this decision-making process? To answer these questions, we analyze data from 56 in-depth qualitative descriptions of recent risky and non-risky sexual episodes from 28 women who were currently living in shelters in Los Angeles County.

Background and Framework

Acquired immunodeficiency syndrome (AIDS) is among the leading causes of death for women of reproductive age (Centers for Disease Control and Prevention, 2007). HIV/AIDS disproportionately affects African American women (Heron, 2007), who also comprise the largest segment of the female homeless population in Los Angeles County, the setting of this study (Los Angeles Homeless Services Authority, 2006; Wenzel et al., 2007). The leading transmission route for HIV infection among female adults and adolescents is high-risk heterosexual contact, comprising 46 percent of reported cases (Centers for Disease Control and Prevention, 2007). Women additionally are at higher risk of contracting HIV through heterosexual intercourse than are men (Centers for Disease Control and Prevention, 2007), partly because women have a higher probability of encountering infected partners than men (Tortu et al., 2000). Furthermore, impoverished women, particularly those who are homeless, are vulnerable to HIV in part because of their greater engagement in high-risk sexual behavior and substance use (McDonnell et al., 2003; Nyamathi, 1994; Sikkema et al., 1996). Because homeless women face greater risk, and substance use and sexual behavior are activities that occur in a context involving other individuals, it is important that research attends to this context in an effort to better understand and ultimately reduce homeless women’s risk for HIV infection. In this study, we define sexual risk behavior in terms of the use or non-use of male condoms since use of male condoms remains the most effective method against the spread of HIV (Seal & Ehrhardt, 2004) and heterosexual intercourse is the primary means of HIV transmission for women.

Although a number of studies have found statistically significant associations between risky sexual behaviors and alcohol and drug use, the complexities of this association are not well understood and have been understudied, particularly among impoverished women (Sikkema et al., 1996). Alcohol use may be an important risk factor for HIV transmission through its association with high-risk sexual behavior (National Institute on Alcohol Abuse and Alcoholism, 2002). Alcohol misuse may contribute to high-risk sexual behavior by reducing behavioral inhibitions and risk perceptions (Cooper, 2002; Fromme et al., 1999; MacDonald et al., 2000). Alcohol may deliberately be used during sex to provide an excuse for engaging in high-risk behavior or to reduce conscious awareness of risk (Dermen et al., 1998; McKirnan et al., 2001; National Institute on Alcohol Abuse and Alcoholism, 2002). Some believe that alcohol enhances sexual arousal and performance and may use alcohol as a prelude to engagement in high-risk sexual behavior (Cooper, 2002; Dermen & Cooper, 2000; Dermen et al., 1998; George et al., 2000).

Alcohol may also undermine women’s efforts to engage in HIV self-protective behaviors (e.g., wearing a condom, abstinence) by increasing women’s susceptibility to violence and coerced sex (Pulerwitz, 2000). The link between women’s alcohol misuse and HIV risk behavior may be due in part to the co-occurrence of drinking with drug use. Crack cocaine use is associated with risky activities including prostitution and sex with multiple partners (Edlin et al., 1994; Schilling et al., 1994) and women crack users may be at equal or greater risk of HIV infection than intravenous drug users (Center for Substance Abuse Treatment (CSAT), 1994; Chaisson et al., 1989; Fullilove et al., 1990).

To understand the role that alcohol and drugs may play in homeless women’s risky sexual behaviors, it is critical to understand the contexts in which sex and alcohol and drug use co-occur. On the one hand, context can refer to the larger social and personal networks in which homeless women are embedded. For example, women may be more likely to engage in risky sexual behaviors if members of their social network misuse alcohol. Such women may be more likely to come into contact with and select high-risk sex partners, misuse alcohol themselves, and be exposed to social norms that increase their health risks. Women also may be more likely to have risky sex if they experience victimization within their social network. Indeed, the general ineffectiveness of interventions to change sexual risk behaviors in adult heterosexual populations has been largely attributed to inadequately addressing social contextual factors (Fisher & Fisher, 2000; Logan et al., 2002; Tortu et al., 2000).

On the other hand, context also can refer to the situations that women find themselves in when engaging in sexual activities. As processes, sexual activities and events unfold over space and time with prior contexts and actions informing subsequent actions and contexts. To understand the full process, we divided sexual events into three stages: pre-intimacy, intimacy, and post-intimacy. Pre-intimacy represents the physical and social context in which people meet and eventually decide to engage in some kind of sexual activity. The intimacy stage refers to the time period in which individuals engage in sexual activities and where decisions about whether or not to use condoms (whether discussed or decided beforehand or at that time) play out. The post-intimacy stage is often a stage of reflection for people and represents the transition back into the larger context of daily life. Although these stages were not theoretically driven, we felt they were a logical and pragmatic way of ensuring we had a comprehensive understanding of a sexual event from start to finish. We can further characterize each stage by what participants do, think and feel, as well as by the social and physical environments in which these activities occur. It is within this experiential context that we focus our investigation. This paper explores how homeless women make decisions about risky sex and what role alcohol and/or drugs play in this process.



We recruited participants during the months of March and April 2006 from community shelters in the Los Angeles County area serving a majority homeless population. Eligible shelters had to be located within an approximate 15-mile radius of downtown Los Angeles, have 8 or more beds, have an average length of stay for women of 12 months or less, and not be a shelter serving exclusively victims of domestic violence or gay and lesbian clientele. These exclusions resulted in a total of 144 eligible shelters. To ensure a diverse sample of women and the sheltered environments they came from, shelters were stratified by region (Metro, San Fernando Valley, and West) and size (small=8–30 beds vs. large= ≥ 31 beds). Two shelters were randomly sampled from each of these six strata for a total of 12 shelters. However, once in the field we were not able to obtain the collaboration of 3 shelters: a large-size San Fernando Valley, a large-size West, and a small-size Metro. Therefore, all interviews were conducted at a total of 9 shelters.

Figure 1 provides a flowchart outlining the process in which participants were screened and interviewed at the shelters. Sixty-six women across the 9 sites were randomly selected from shelter-provided bed lists using a random numbers table. Trained female interviewers gave a brief presentation inviting women to participate in the project. Women who were interested in participating completed informed oral consent and a screener for eligibility. Sixty-four women (97%) completed the screener and 34 (52%) screened eligible. Women were defined as eligible if they were at least 18 years of age, had sex with a male sexual partner within the past 6 months, and were able to complete the interview in English. Of those who were not eligible, this was mainly due to women reporting they had not had sex with a male partner within the last six months (90%) and then a much smaller proportion were not able to complete the interview in English (10%). Women who were eligible and interested in participating in the interview completed informed written consent and participated in the 1-hour interview in a private area of the shelter. Of the 34 women who were eligible, 15 percent did not complete the interview (n=5) either because they did not show up for the interview (n=4) or they refused to partake in the interview after learning about what their participation would involve (n=1). Additional criteria for the analyses presented here was a minimum requirement that women discuss two sexual events within the last six months, which eliminated 1 additional woman. This resulted in a total of 28 women who completed the interview. Approximately 1–5 interviews were completed at each site with the distribution of interviews evenly distributed across geographic region: Metro (32%); San Fernando Valley (32%), and West (36%). Completed interviews were also fairly evenly split between small (46%) and large (54%) sized shelters.

Figure 1
Flowchart of Participant Recruitment

Our final sample of 28 women interviewed ranged in age from 18–63 years old (mean = 37 years; sd = 9.8), reported being homeless from 0–15 years (mean = 3.8 years; sd = 3.8), and indicated 9–16 years of completed education (mean = 12 years; sd = 1.5). Of the 28 women, 13 (46%) were African American, 6 (21%) were Hispanic/Latina, 6 (21%) were non-Hispanic white, and 3 (11%) were multi-racial or mixed race.

Even though both shelters and women were selected randomly, we recognize that our sample might, for example, over-represent women from small shelters. For this reason, the sampling design was such that the obtained sample is not self-weighting and should not be used to generate population prevalence estimates for homeless women who use shelters in Los Angeles.

The sample, however, is adequate for exploratory studies like this one where our objective is to identify the range of behaviors and beliefs among women and to start to identify potential patterns that could be tested in future studies employing larger and more representative samples. Although there are no strict rules for determining sample size for such exploratory qualitative investigations, such studies typically include 30 or fewer respondents per group of interest (e.g., sexual behavior variants) (Patton, 1990). Some qualitative researchers have recommended using as few as 4 or 5 respondents per group to understand the essence of experience (Morse, 1994).

Data Collection

We used in-person, semi-structured interviews to elicit information on two sexual events for each respondent resulting in a total of 56 events (28 women × 2 events each). To be able to gather information about a diverse set of situations we used an event-level study design and asked participants about: a) their most recent sexual event; and b) their second most recent sexual event that differed from the first event. We asked them to pick a second event that differed in terms of condom use or, if that did not change during the last six months (i.e., always used a condom or never used a condom), an event that differed by alcohol and/or drug use (i.e., an event that did not involve alcohol or drugs if the first did and vice versa). This is a comparative case-study design that combined the use of a ‘between-subjects’ and a ‘within-subjects’ model. The ‘between-subjects’ design asks respondents to report on their last risky episode to accurately identify social context-level variables associated with risky behavior (Leigh et al., 2008; Tortu et al., 2000; Vanable et al., 2004). However, this method does not eliminate the possibility of confounding individual characteristics. For example, a risk taker may be more likely to engage in substance use on any given occasion and be more likely to engage in unprotected intercourse on any given occasion. In contrast, a ‘within-subjects’ design involves each respondent reporting on two events, one that includes risky behavior and one that does not, and thus holds constant relevant individual differences such as predisposition towards risk taking (Bailey et al., 1998; Testa & Collins, 1997). With this combined design of asking about two recent but distinct events we aimed to understand both the contexts in which women engage in sex as well as to identify event-level characteristics that may influence risk behaviors, in particular inconsistent condom use and consumption of alcohol and drugs (Leigh & Stall, 1993).

For each event, we began by asking women to describe the experience in their own words and in as much detail as possible. After this “grand tour” question, we asked women to describe their relationship with their partner (e.g., Is he a steady partner like a boyfriend or husband, a casual partner that you have had sex with once in a while, or a partner you might have sex with just for things you need, like a place to stay, food, money or something like that?) and then used a topic-by-stage framework to probe for more specifics (see Appendix for complete protocol). The topics for the framework included: context (e.g., when and where did you meet up, were you familiar with that place, who was there); behavior (e.g., what happened, was the meeting planned, who made the first move, what sexual activities occurred); thoughts and decisions (e.g., what were you thinking, how did you decide to have sex, did you think you would have sex); feelings (e.g., what were you feeling, what was your partner feeling); alcohol and drug use (e.g., what and how much, what role did it play in your decision to become intimate); condom use (e.g., did you use condoms, who had the condoms, did you have a discussion about using a condom); and comparisons (e.g., how typical is this, how similar or different is this from other times you have met up for sex). Dividing the event into stages allowed us to utilize three fundamental time periods: (a) pre-intimacy or the period when the couple met and eventually decided to become intimate; (b) intimacy or the period during which sexual activity occurred; and (c) post-intimacy or the period after intimacy.

Stage-By-Topic Interview Protocol

The interview protocol displayed this topic-by-stage framework as a grid with checkboxes next to each question which allowed the interviewer to ensure all topics and time periods were covered with each interviewee. We asked respondents open-ended questions before we asked closed-ended questions so as not to bias respondent answers as well as to be able to explore new leads and to generate richer and more detailed personal narratives (Bernard, 2006; Spradley, 1979). Since we were particularly interested in decision-making with regards to risky sex, we probed specifically about condom use and alcohol and drug use as their own topic areas. Participants were given an incentive of $20 for their participation in the interview. Interviews were audio-recorded and transcribed. Transcripts were managed and coded using ATLAS.ti (Muhr, 1991). Human subject protections and data safeguarding procedures were approved by RAND’s Human Subject Protection Committee.


We based our analysis on a comparative case-study design (with sexual episodes as the units of analysis) to qualitatively isolate general factors associated with sexual risk-taking and alcohol and drug use. To identify themes, we utilized a staged technique described by Lincoln and Guba (1985) and elaborated on by Ryan and Bernard (2003). First, we used text management software (ATLAS.ti) to mark contiguous blocks of transcript text that pertained to the major topical domains of interest outlined (context, behavior, thoughts and decisions, feelings, alcohol and drug use, and condom use) and the three time stages of the event (pre-intimacy, intimacy, and post-intimacy). We then pulled out all text associated with a particular domain and divided the texts into shorter “quotes” that expressed a single general idea and that could be read as independent statements. After printing the quotes on slips of paper, we next spread them out on a large table. Four team members studied the quotes and engaged in a discussion about how to best sort them into piles based on their thematic similarities. Eventually, these team members came to a consensus about what thematic categories to include and which quotes were affiliated with each. We then named each thematic category and developed an explicit codebook to describe each (Crabtree & Miller, 1992; MacQueen et al., 1998; Miles & Huberman, 1994; Willms et al., 1990). In the next step, a two-person team matched each quote in a domain with a specific subcategory. In cases of disagreement or confusion about how to code a particular quote, we first examined the larger context within which the quote came. If it was still unclear, we asked the original field interviewers if they could shed light on particular instances. If the issue was still not resolved by consensus then the decision fell to the project leader. We then examined the degree to which these themes were distributed across stages, women, and events.

To ensure that we had accurately assigned quotes to themes, we also conducted an independent reliability test. For each of the main domains, we randomly selected a subsample of quotes and had a fifth team member who had not been involved in the original sorting or coding tasks match them with our codebook. We calculated a Kappa statistic for each domain (Cohen, 1968). With one exception, all Kappa statistics were above .85 – an acceptable level by most standards (Brennan, 1992).


To best describe how sexual events unfold over time, we have broken the events into three stages: pre-intimacy, intimacy, and post-intimacy. In the pre-intimacy stage we describe how, where, and under what circumstances people met before engaging in sexual activities, and the kinds of relationships and emotional connections women report having with their partners. During the intimacy stage, we report on what kinds of sexual activities women engaged in, and what they thought and felt during this phase of the encounter. In the post-intimacy stage, we examine women’s thoughts and decisions and compare the described sexual events to what they report as their typical behavior. Finally, we examine how women made decisions regarding condom use and what role alcohol and drugs played across the three stages.



In general, sexual encounters were slightly more planned than not. In over one-half of the encounters (30 of 56 events), women reported that prior to meeting with their partners they anticipated that they would engage in sexual activities. In some cases (17 of 56 events), sexual activities were explicitly recognized as one of the principle reasons for getting together. As one women described, “We talked about it and we planned on having our little weekend together. Our weekends are pretty typical. We go to get something to eat at the market. We get a room with a microwave and a refrigerator, so we can put all our goodies in. And then we eat, have sex, or have sex and eat - whichever way it goes.”

In other cases, planning for sex was not made explicit but was often assumed or at least anticipated as a likely outcome. For instance, we found multiple examples where previous sexual partners (usually ex-boyfriends, ex-spouses, or estranged husbands) would initiate contact and suggest meeting. In other instances, it was the woman who initiated contact. In such encounters, the purpose for the meeting was typically nonsexual. For example, one woman went to pick up her children’s backpacks from an ex-lover’s house and ended up staying and having sex. Only in a fifth of the encounters (10 of 56 events) was sex unplanned in that either one or both of the partners did not have an expectation of sex and the physical intimacy progressed “spontaneously.”


Most of the sexual encounters took place either in a hotel/motel (20 of 56 events) or at a male partner’s home (14 of 56 events) with a few exceptions such as in a vehicle, at a river wash, and at a shelter. Typically, people met in the same place they had sex. Approximately one-fifth of the respondents (11 of 56 events), however, mentioned meeting up at a particular locale, and then moving to a different venue for sex. Women mentioned meeting their partner at a friend’s house, in front of their own shelter, at his work, at a Narcotics Anonymous meeting or other pre-set location, and then moving to a final destination for the sexual activity. In such encounters, the couple often met, ate a meal and/or ran errands, and finally ended their time together with sexual activity.

Relationships and emotional closeness

Across these sexual events, the homeless women reported a wide range of different sex partners. Of the 28 women interviewed and the 56 sexual events described, all respondents indicated that they knew their partners to varying degrees before the event. None of the women reported having anonymous sex or “one-night stands” with strangers. Even with respect to the sex trade, respondents indicated that they had recently only had sex with men they “dated” regularly or previously knew.

Perhaps not surprisingly, women often characterized sexual events in terms of the emotional intensity they felt toward they partner and the duration of the relationship. In 35 of the 56 sex events, women described their relationship with their partner as “serious,” whereas in 21 events they described their relationship as “casual.” The serious relationships were characterized as involving deeper emotional connections and were typically associated with words and phrases such as: “long term,” “husband,” “boyfriend,” “live together,” “love,” “kids,” “close,” “commitment,” “courting/dating,” and “supportive.” In 14 of the 35 events described as being with serious partners, women identified their partner as “husband” or “common law husband.”

Serious relationships could be quite intense and relatively stable over longer periods of time. For example, one woman reported, “We love each other dearly – very, very, very much. And we feel like we’re kind of common-law husband and wife. So it’s like a husband-wife relationship.” Another woman described the evolution of her relationship this way, “I call him every day, because he started out as my friend, became my best friend, and then I slept with him. And after I slept with him, I just fell for him. I mean, I didn’t kiss him. I never would kiss him on the mouth. And then, one day, I kissed him, and after that, I was falling for him—the kiss of death.” Even rocky relationships could be filled with emotion. As one woman noted: “We’ve had our ups and downs. I just really love him with all my heart. And since the day that I met him, I knew that that’s who I wanted to be with, even though we did drugs together. I still felt that I want to be with this guy. I don’t know what I would do without him.”

Serious relationships ranged in duration from several months to several years and, in a few cases, over twenty years. Several women had daily face-to-face or telephone interaction with their serious partners often depending on their shelter program and what resources they had access to. In other serious relationships, the women had little or no contact with their partners, some of whom were incarcerated or in other shelter programs, but they considered the relationship active and serious.

When women referred to casual sex, they were typically referring to sex with a relatively new partner or within a relationship that only lasted for a short period. The emotional intensity of such casual relationships, however, ranged from romantic to purely economic, with sex with friends or “booty calls” falling between these two extremes. The romantic casual encounters occurred with men who might be potential boyfriends, whereas the “booty calls” were with male friends who respondents did not feel attached to emotionally. One woman described a “booty call” this way, “There ain’t no relationship. It’s just that I’ve known him for a long time, and he's just a sex partner. He’s just a good friend that I socialize with sometimes.” Although she may have known him for a long time as a sex partner, this respondent did not define a lengthy sexual relationship as a serious relationship. Another woman said, “I thought maybe we would just kind of hang out…I fell into it…I mean, there’s no regrets. But, he isn’t somebody that I really care about.” In contrast, casual sex sometimes also included a sense of closeness that was supplemented with sexual activity. For instance, one woman described her relationship thus, “We watched TV together. It was more of a friendship, with sex, than just sex, you know?” In other cases, previously strong relationships were apt to become less emotionally intense over time, but not lose some of their physical intimacy. For example, another woman described her relationship by saying, “He was…somewhat like a boyfriend, at one time. Now it’s more like casual.”

Women also described activities that included exchanges of sex for money, food, shelter, alcohol and/or drugs, or other material things. This could occur both as part of the sex trade or in less formal situations (e.g., a place to stay or dinner). As one woman put it, “This is like I just fucked him for money.” Even a lesbian woman admitted to engaging is such heterosexual activities: “…it’s a hurtful feeling. You know? I’m not happy at it at all, because I never want to do it. But when he calls, I go, because I know there’s money.” In approximately one-fifth of the episodes (11 of 56 events), respondents reported exchanging sex for money, drugs, or material things. Of the 11 events that were described as sex trade, 5 events were described as being in a “serious” relationship with their partner.



In describing their sexual behavior, in almost one-half of events (24 of 56 events) respondents reported that the meeting took place over an extended period of time (e.g., a couple of days) or overnight. It is not surprising because, as noted previously, many couples were in serious relationships and rented hotel rooms for their time together that one respondent referred to as a “mini-vacation.” These longer sex encounters were contrasted with “quickies” mentioned in 10 percent of the events (5 of 56 events) in which a sex encounter was of short time duration, typically less than an hour. The other half of the meetings took place within a duration of more than an hour but less than an overnight stay (27 of 56 events).

All of the women reported having vaginal intercourse during their sexual encounter; additionally, one-half also mentioned oral sex (with slightly more fellatio (36%) than cunnilingus (30%)), and a few mentioned masturbation (9%). Kissing, fondling, massaging, cuddling, and undressing one another were also behaviors mentioned during the period of intimacy. Only one woman in the sample reported that she engaged in anal sex. Most women, when asked about anal sex, said that they did not engage in anal sex because it was “painful.”


During the intimacy period, respondents were more likely to express positive feelings such as being “happy,” “aroused,” and “relaxed” (42 of 56 events) than they were to express negative feelings such as “wanted it to end,” “anxious,” and “painful” (28 of 56). Only a handful of respondents (9 of 56 events) were ambiguous about what they felt or did not recollect their feelings during sex. A few women mentioned both positive and negative feelings about partners, mainly expressing sadness over long-term relationships that were not working out. In terms of the male partner’s feelings, as perceived by the women, the bulk of the comments were about his feelings of happiness or being “horny” with a few comments about missing the woman and not wanting their time together to end.


To more thoroughly understand how women experienced sexual events and their thoughts about the choices they made during the process, we also asked respondents to describe how they felt after having sex. Respondents expressed a range of emotions and thoughts during the post-intimacy stage. The women in serious relationships were happy to have spent time with their loved ones, but wished they could have stayed and felt sadness when parting. As one woman put it, “We sometimes, after we have sex, we just lay there and just hold each other, and tell each other how much we love each other, and sometimes it gets emotional and we’ll cry, you know? And it’s really how deeply we feel about each other.” Another woman echoed similar sentiments, noting that “after he came, he stayed inside of me, and we kind of hugged each other…. We just held each other. I mean, there was no real verbal communication and it was just more spiritual and understanding that we were on the same page in terms of how we felt about each other.” A few respondents mentioned feeling good, satisfied, and relieved with the release of tension the sex event brought for them.

In other cases, women expressed a relative indifference. “I don’t really know that I felt a whole lot of anything, because I was used to doing that. I’ve been doing heroin for like 20 years, so I’ve been doing that type of stuff for 20 years, you know to support my habit. If anything I felt like, you know job well done, or something. I don’t know. Just like, good girl, you know? Because it was fast, it was quick, and he was happy. I was happy.” Another woman put it this way: “I felt all right. Oh, good, to me. Just sex. I mean…it wasn’t nothing spectacular about him, or nothing special. It wasn’t like I was in love with him, or felt any emotions about it. Just a…sex like an act being performed. Just something to do to provide [money] for me at that time.”

Other women discussed the regret, shame, and guilt they faced when the meeting was over. For instance, one woman described her post-intimacy feelings thus: “I felt kind of relieved, because I hadn’t had sex in like ten months. I felt bad, because I was, like, oh, God, I just had sex, oh my God. Felt kind of bad…Because it was, because we’re not married. We’re not even dating each other. We were just pen pals. You know, I hardly, hardly knew him but just we got drunk.” Another woman expressed similar misgivings: “I was nervous, because it was the first time, and I was cheating technically. What am I doing to myself was running through my head the whole time. What am I getting myself into? You know you wear your heart on your sleeve [I said to myself]. Why? What are you getting yourself into?


Of the 56 events, condoms were used in 19 events (34%) and not used in 37 events (66%). During several encounters, couples began by using condoms, but decided to remove the condom at some point during intercourse. We counted these as non-condom events. Although some women reported that they always used condoms, particularly in the face of contracting an STD, other women provided a wide range of reasons for using and not using condoms. Some women said they did not use condoms because they were in a long-term relationship and/or they and their partner had done STD testing together. Others mentioned not using condoms because they “trusted” or “felt safe with” their partner. Still others mentioned not thinking clearly during the “heat of the moment,” not liking the physical feeling of using a condom, feeling that condom use reflected poorly on their own self perceptions, being high or intoxicated, being infertile or menstruating, feeling uncomfortable asking their partners to wear a condom, or because the man always “pulls out” before ejaculating. We examine some of the more salient rationales below.

Serious versus casual relationships

One of the most relevant factors mentioned by women was the strength of their relationship with their partners. We found that condoms were used in only 23 percent of the sexual events (8 of 35 events) where women described the relationship with her partner as being “serious.” In contrast, condoms were used in 52 percent of the sexual events (11 of 21 events) where the woman described the relationship with her partner as being “casual.” For some, it was just a matter of time. One woman described her decision making in this way: “When we first met, we would start out using a condom, but since we’ve been together…we decided not to use it, because it was just me and him…no other kind of sex with nobody else, so, we didn’t have to use condoms.” For others, the decision involved time, testing, and trust. For example, a woman told us, “We were both tested. Once we decided that we were going to live together and be together. And I just trusted-just trusted him, and trusted myself.” For women already in long-term relationships, even being asked by the interviewer about condom use seemed strange and somewhat uncomfortable for the women. When we asked one woman about using condoms with her long-time partner, she responded: “Wow, he would look at me crazy now because it’s been like years, and like why are you bringing it up? So, I think he would be uncomfortable, because he’d probably-like, why you want to use a condom? Are you having other sexual partners? Are you cheating on me? And we don’t want to go through that.”

Trust and safety

Women’s perceptions of trust and safety with their partner also influenced decisions about condom use for both serious and casual relationships. Clearly, trust and safety are part of what makes up a long-term, serious relationship. For example, one woman explained why she did not use condoms with her long-time partner this way: “Because we’ve been with each other for so long. I don’t know. I won’t use a condom with him because I feel safe with him. I know he isn’t with other people. I could trust him.” But perceptions of trust and safety also play an important role for more casual partners as well. For example, one woman who did not use a condom said she was worried “because there’s always that line that we might be [infected] but, I just felt like he wasn’t, you know, I felt safe. But you’re never 100 percent sure of who your partner is with. I mean, because you can’t be with him 24 hours a day there’s always that possibility that, hey, he might be sleeping around. But, I felt pretty confident.”

Issues of trust were often intertwined with other considerations as part of the decision-making process. For instance, one woman described her thought process as such: “I trusted him, even though I don’t know him. I just trusted that it would be okay. I’m still worried. But … at the time, I didn’t, I really didn’t. I just wanted to do it and that was the main thing, trying to please myself, my desire.” When we asked another woman why she did not use a condom, she replied, “It feels safe with him. And plus, I can’t get pregnant. I’ve had my tubes cut, and I feel really safe with him, so I can’t get pregnant. I feel safe with him. I trust him enough that he won’t give me anything. And that’s the only reason why.” A third woman explained why she did not use a condom: “I guess because he felt safe, because I had been in jail for nine months. So he didn’t-he just felt like I hadn’t been sleeping around. So he felt safe. Although, I’m not sure if I felt safe with him. Now that I think about it. But at the time, I didn’t even think about it, because I was drunk.”

Unless there were other extenuating circumstances, lack of trust often led women to insist on using a condom. For example, a woman described taking a condom out of her purse and asking someone she had sex with on numerous previous occasions to use it: “He got upset and said, ‘Well, why I got to use this? You ain’t never asked me to use a condom.’ I said, ‘Well, I don’t trust you. I don’t know where you’ve been.’” Another woman was even more adamant: “Oh, I told him that I wasn’t going to do anything with him unless he had a condom. And he said, ‘Well, why? I just got tested like two months ago.’ I said, ‘Because I know you’ve been, you know, seeing this person and seeing that person.’ And he was like on a little mission to try to do everybody up there before he left, and I wasn’t trying to take chances with him. I don’t know exactly who it was but, you know, it’s better safe than sorry. Plus at that point I wasnt on any kind of birth control or anything, as well. And it just would be unfair to my kids if I got pregnant at this time.”

Assessments of trust also seemed to be made on the fly, particularly for partners they had not known for long periods of time. One woman described thinking about using a condom after her partner had mentioned that he slept in a bed with a female friend of his: “He said they just laid there, but I was kind of like, I think I better use a condom. Then I was like, well no, I said, no, I trust him. He say he didn’t, he didn’t. So I said, I ain’t going to bust out the condom, because he’s going to think I don’t trust him.” In other cases, women seemed to base the assessments of safety on relative comparisons. One woman noted that the men she knows who use a lot of alcohol and drugs are “kind of losers [with regards to sex so] when I meet a man that’s got a working job, or appears to have a working job, and they’re clean or have kids with them or something I feel safer. “

The role that perceptions of trust and safety play in condom use decisions appears to be somewhat different with casual partners compared with serious partners. For casual partners, Table 1 shows that 70 percent of women who did not use condoms spontaneously mentioned that they trusted or felt safe with their partner. In contrast, only 18 percent of the women who used condoms spontaneously mentioned these attributes in their descriptions (2-Tailed Fisher’s Exact, p=0.03). We suspect that for most women, casual relationships imply lack of trust and safety. When these issues are brought up unprompted they are likely to indicate that the woman has more trust in her partner and therefore is less likely to use condoms. In serious relationships, women were equally likely to spontaneously mention trust and safety but this was less likely to be related to not using condoms rather than using condoms. In such cases, less than half of the women who did not use condoms (44%) spontaneously mentioned trust or safety, and slightly more than half who did use condoms (63%) spontaneously mentioned such attributes (2-Tailed Fisher’s Exact, p=.44). We suspect that for most women, serious relationships imply trust and safety. When these topics are brought up unsolicited, they are more likely to indicate a woman’s concerns about her partner’s loyalty and sexual health.

Table 1
Relationship between women’s spontaneous mention of trust or safety and condom use controlling for type of partner

Perceptions of others and self

Some of the women in our sample held negative impressions of men who had condoms readily available. For example, several women made statements such as, “he must have been a pig” or “he was a health fanatic” to have condoms so readily at hand.

Conversely, several other women felt that the man’s use of a condom was an insult to her reputation. One woman said, “I think he was scared not to use a condom. I think he thought I was a whore or that I slept with everybody, and he wanted to use it. That’s what I think.” Another woman reported that she never uses condoms and “I get mad if somebody ask me if I need any.” Another woman said, “When I use condom, I feel like I’m not worth having sex with, or like, you think I have something. If you are in a relation with somebody and you touch that person and you love that person, I don’t think it’s necessary to use them.”


Condom use involves a two-step process – first, having condoms available at the time of sex and second, deciding to use them. In our sample of events, women reported that condoms were available 39 percent of the time (22 of 56 events), but were only used 30 percent of the time (17 of 56 events). Most of the cases in which condoms were not available could be attributed to more serious relationships in which the decision to stop using condoms had been made sometime in the past. Interestingly, we had no cases in which women said they failed to use a condom because none were available. In fact, in one case a woman asked her partner if he had a condom and when he did not, he ran across the street, bought a condom, and returned 10 minutes later with it. Men appeared to have slightly more control over condom use. When the man was the only one who had a condom (n=11), it was used 82 percent of the time (9 of 11 events), and when the woman was the only one who had a condom (n=8) it was used only 63 percent of the time (5 of 8 events). When both partners had condoms (n=3), they always used them. For the five events in which condoms were available but were not used, two factors appeared to be primarily responsible. In two of the events, the decision not to use a condom was driven by physical feelings, including because the couple was “too hot and passionate” or because both partners wanted to “feel” each other without the condom. In the other three events, the decision not to use a condom came from the woman worrying about the social dynamics between herself and her partner rather than worrying about any individual health risks she could be exposing herself to.


When having sex in unfamiliar venues, women reported using condoms 37 percent of the time (14 of 38 events) compared to using condoms 28 percent of the time (5 of 18 events) in familiar venues. Table 2 shows that the effect of venue differs by type of partner. For instance, in unfamiliar venues with casual partners, women were equally likely to use a condom (50%) or not (50%); while in unfamiliar venues with serious partners, women were more likely to use condoms (29%) than not (9%) though the relationship is not statistically strong (2-Tailed Fisher’s Exact, p=.39).

Table 2
Relationship between familiarity of venue and condom use controlling for type of partner

Alcohol and Drug Use

Out of the 56 events, 32 (57%) did not involve any alcohol or drugs, whereas 24 events (43%) involved either alcohol (15 events), drugs (14 events), or both alcohol and drugs (5 events). Of the 24 events where alcohol and/or drugs were used, 50 percent of the time both partners used substances, 25 percent of the time the woman used alone, and 25 percent of the time the man used alone. Among the 15 events involving alcohol use, beer and wine were used in 10 events and hard liquor was used in 5 events. In the 14 events where drug use was reported, cocaine (crack and powder) was most widely used (7 events) followed by marijuana (4 events), heroin (3 events), and ecstasy and methamphetamine (1 event each). Consumption of substances in small amounts was reported two-thirds of the time, including amounts such as a single beer, a single joint, or a dime bag of pot. The other one-third of the time, there were relatively large amounts of substances consumed, such as a fifth of alcohol, “$200 worth of drugs,” or a gram of crack.

The role of alcohol and drugs in deciding to have sex

The role of alcohol and drugs in the decision to have sex appeared to strongly depend on the main reason the encounter was initiated. We identified four primary reasons women met their partner: (a) to acquire drugs; (b) to obtain resources (e.g., money, shelter, food, etc.); (c) to be intimate or emotionally close; and (d) to fulfill physical and sexual desires.

In the events where the primary goal was to get high, sex was often used in exchange for drugs. As women explained, in these settings the sex was mechanical and the situation was uncomfortable. One woman described such an encounter this way: “After a while I’m getting high, and I don’t think about nothing but getting high. And it irritates the fuck out of me. And he’s the only one. I hate being around him, but like I said, I know I get money and then I can smoke all day. When he’s getting high and I’m at his house, I’m very, very uncomfortable, but one more time, I say, that’s drugs and money. Free drugs and money….You know, I wasn’t happy. And I remember I was there and I just threw him off of me. I put my clothes on, but I sat there, because I knew he still had drugs and money. ” Another woman described the mechanical aspects of her encounter: “Him kissing on me, and I was like, UGH because he likes to kiss, and I am really not a kisser. Forget the kissing. Let’s go straight to the sex because I wanted it to be over, so I could smoke my dope. And this is so basic, that it’s really a shame, because there’s no foreplay, there’s no talking and we just start having sex…And because I don’t give it to him often, it took like two minutes and it was over. I think it took longer for me to wash up and get dressed to go back to the dope.”

In events where the primary goal was to obtain resources, women sometimes used alcohol and drugs as a way to overcome some of their discomfort. As one woman told us, if she had not been high “I would have never done that. If I didn’t need money for whatever, somewhere to sleep, or dope, or whatever, I would have never done that”. Another woman said, “I don’t believe that I would have had sex with him if there weren’t drugs involved…I pushed him off a few times, and he was fine with that and then, later, I consented to it. I don’t even know why I did. I was tired and basically just wanted to get it over with.”

In the events where the primary goal was for emotional intimacy or to fulfill sexual desires, alcohol and drugs did not appear to be a major influence in whether women decided to engage in sex or not. Although alcohol and drugs were often present in such events, the decision to have sex seemed to be driven by other factors. For example, the desire for emotional closeness was quite powerful for some women. One woman described it this way, “I was feeling like an intense, emotional need for him. Like I needed him, and I wanted him. And it was more loving, and more used to him. So just that feeling of old past. I just wanted that love back. How it used to be.” Such an emotional connection was often contextual. For instance, another woman noted that “…this time we weren’t sure if we were ever going to see each other again, and we really do deeply love each other …and so…it was very intimate lovemaking.” Such feelings were clearly in contrast to those encounters where emotional closeness was absent, for example when women were engaging in sex for pure physical pleasure: “There was no intimacy there. Nothing. Nuh-uh.” As another woman said, “I was feeling horny and excited, and I wanted to do it, and I wanted to have an orgasm, and stuff. And that was my purpose…and I wanted it to be good.”

Role of alcohol and drugs in deciding to use condoms

In general, condoms were used in 29 percent of events (7 of 24) involving alcohol or drug use and 38 percent of events (12 of 32) not involving alcohol and drug use. The use of alcohol, drugs, and condoms during their sexual meetings were the topics the women discussed with clear hindsight, as noted by this respondent: "I honestly think if I wasn’t high, if I was more clear-minded, I would be thinking about condoms, and not getting diseases, AIDS, whatever. But when you’re in the moment-you’re partying, you’re in the moment, you’re not thinking of that shit. Which is very irresponsible, but that’s how it is." Women worried about the threat of STDs, of getting pregnant, and of hurting themselves emotionally, after the intercourse took place. A few respondents remembered calculating these risks while engaged in intercourse without condoms. Even one respondent who said she thought alcohol did not affect her decision-making offered that it made her more “loose” and ready for anything.

At this aggregate level, the relationship between drugs and alcohol and condom use is not statistically significant (2-Tailed Fisher’s Exact, p=1.0) – women are equally likely to use a condom whether or not they used drugs or they consumed alcohol. The relationship strengthens considerably, however, when we examine drug and alcohol use with serious and casual partners. Table 3 shows that in sexual events with casual partners, drug use is associated with increases in condom use; while in events with serious partners, drug use is associated with decreases in condom use. In sexual events with casual partners, condoms were used in 83 percent of the events (5 of 6) involving drug use, but only 40 percent of the events (6 of 15) not involving drug use (2-Tailed Fisher’s Exact, p=.15). The relationships are the opposite for serious partners. In sexual events with serious partners, condoms were never used (0 of 8 events) in events involving drug use, but were used in 30 percent of the events (8 of 27) not involving drug use (2-Tailed Fisher’s Exact, p=.15). Table 4 shows a similar, although a statistically non-significant pattern for the use of alcohol. In events with causal partners, condoms were used 57 percent of the time (4 of 7) when alcohol was consumed compared with 50 percent of the time (7 of 14) when alcohol was not used. In events with serious partners, condoms were used 13 percent of the time (1 of 8) when alcohol was consumed compared to 26 percent of the time (7 of 27) when alcohol was not consumed (2-Tailed Fisher’s Exact, p=.65).

Table 3
Relationship between drugs and condom use controlling for type of partner
Table 4
Relationship between alcohol and condom use controlling for type of partner

Additionally, a group of 8 women, representing approximately 30 percent of respondents, reported that their use of substances (alcohol and/or drugs) was so common that they felt it did not play a role in their decision-making around sex and condom use. This behavior is exemplified in this comment: "I don’t think drugs and alcohol really had a role, because I’m so used to it. I do [drugs and alcohol] every day so I actually function very normally."


These detailed descriptions of 56 sexual events are illuminating in that they highlight the considerable diversity and complexity in the sexual relationships of homeless women and provide additional insight into the growing literature on the factors that influence safer sex practices among impoverished women. We summarize our results below.

First, we found that homeless women had a far broader range of sexual relationships than perhaps is commonly presumed. They reported trading sex for money and drugs with “Johns;” having sex with acquaintances, friends, boyfriends, and ex-lovers; and maintaining long-term, stable relationships with husbands and common-law partners. Interestingly, in none of the events did women describe having sex with strangers.

Second, we were struck by how often this sample of homeless women described their emotional attachment toward some of their sexual partners. Not only did they describe strong emotional bonds with their long-term partners, but they were clearly emotionally attached to many of the friends and ex-lovers with whom they had sex. It may be the case that some of these women are seeking to fulfill a need for emotional attachment when they have sex with some of their more casual partners. Seeking emotional attachment may come at a price for these women if it discourages them from using condoms with partners who they do not know well or with whom they do not have an exclusive relationship.

Third, our results highlight the fact that all encounters are not equal. Although we only looked at encounters where sex occurred, the reasons for meeting up in the first place varied dramatically. Sexual encounters where the primary purpose of the encounter was to acquire drugs or resources such as money, food, or shelter were quite different from encounters where the primary purpose was for emotional intimacy or for physical and sexual desires. These differences in turn affect the factors that homeless women consider in choosing to have sex or use a condom, including the role that alcohol and drugs may play in such decisions. Decisions about who to meet and for what purposes put women at different kinds of risk. For example, women who meet primarily for drugs or for resources are in a particularly vulnerable position.

Fourth, our 56 interviews suggest that homeless women are relying on a wider array of types of people to fill their needs for not only resources (and sometimes drugs), but also for emotional intimacy and physical desires. What seems to put homeless women at most risk is that many of them appear to profess a level of trust toward their casual partners that is typically reserved for long-term and emotionally close partners (e.g., boyfriends, husbands, and common-law partners). That is, women in our sample appear to grant some ex-husbands and ex-lovers, friends, acquaintances, and even some “Johns-they-regularly-date” with similar kinds of trust and safety that they would afford an established boyfriend or husband. In contrast to husbands and boyfriends, women may have little knowledge of or influence over the sexual activities of ex-husbands, ex-lovers, friends, and acquaintances. Possibly, it is these trusted casual partners who represent the most risk to the women.

Fifth, our results suggest that relationship commitment and involvement set the context for other factors that may influence condom use. Therefore, it may be worthwhile to look at condom use decision-making separately with serious and casual partners. Consistent with other research, we found that women at risk for HIV infection are less likely to practice safer sex if they report higher levels of relationship commitment (Castañeda, 2000) and involvement (Morrill et al., 1996) and were less likely to use condoms with longer term or steady partners compared with casual partners (Marin et al., 1998; St. Lawrence et al., 1998; Tortu et al., 2000; Wingood & DiClemente, 1998). Our results are also in line with findings from recent studies by Tucker et al. (2007) and Harvey et al. (2006) that found relationship commitment predicted more frequent engagement in unprotected sex with a partner, even after controlling for the type of relationship (primary vs. casual). Our results suggest that in terms of safe sex, factors like drugs and alcohol as well as women’s perceptions of trust and safety play out within the complex context of social relationships.

One of our most striking findings is that condom use was more likely when drug use occurred prior to sex with casual partners, but less likely when drug use occurred prior to sex with serious partners. A non-significant trend in the same direction was found for alcohol use prior to sex with casual versus serious partners. Although little attention has focused on the possibility that substance use might be associated with less risky sex under certain conditions, this result is consistent with a recent diary study by Leigh et al. (2008) showing that drinking was associated with increased condom use with casual partners among women. Alcohol myopia theory (Steele and Josephs, 1990; MacDonald et al., 2000; George et al., 2007; Davis et al., 2007) may provide one explanation of our findings. This theory proposes that alcohol intoxication (and we suggest possibly drug use) restricts attention capacity so that people are highly influenced by the most salient cues in their environment. For sex with casual partners, this may be related to sexual risk. Alternatively, these findings could be due to differences in how much alcohol and drugs were used with casual and serious partners. We fully recognize that such explanations are speculative. Unfortunately, our limited sample size makes it difficult to untangle the interaction between partner type, substance use, and condom use and the probable confounding with sexual motivation and risk perception. For example, some women described situations in which they had sex to obtain drugs or money that they might not otherwise have had. In these situations, their casual partners may be known to be particularly risky and the women may be more determined to use condoms. Perhaps they anticipated and planned for the events and brought condoms. The decision to use a condom may be made, and the sex may be over, before the intoxication occurs. In such circumstances, it may be impossible with these data to disentangle a complex and confounded interplay of partner choice, sexual motivation, trust, safety, substance effects, and condom use decision timing.

Our data also suggest that how much a woman trusts or feels safe with her partner is a driving force in predicting condom use and is particularly powerful when it comes to making choices with more casual partners. As mentioned above, we suspect that some homeless women have expanded the range of partners they trust or feel safe with to include casual partners such as ex-boyfriends and ex-spouses; that is, people whose sexual activities are less well known and over whom they have less control or influence. As Misovish et al. (1997) note, using or not using condoms symbolizes trust, emotional intimacy, and commitment within the relationship. Unlike many other studies that rely on surveys, the data from our detailed interviews allows us to better understand how such symbolic linkages play out in real-life decision-making processes – often in complex ways. For instance, we found events where women reported not using condoms because they felt emotionally close to their partners and trusted them to keep them safe, as well as events where women reported not asking their more serious partners to wear a condom because this would indicate that they were not committed to the relationship.

Our work also adds to the discussion over the role that gender-based power dynamics may play in safe sex practices. Some researchers suggest that relationship power is a key contextual variable that shapes women’s ability to engage their partners in HIV-related protective behaviors (Amaro & Raj, 2000). Others suggest that power dynamics by themselves are not sufficient to explain risk among women and girls (Gutiérrez et al., 2000). Our data suggest that the role of power dynamics varies from case to case. We find examples of women who explicitly (and sometimes quite adamantly) take control of decisions about condoms, and we find examples of women who (for a variety of reasons) appear to relinquish control of this process. Our results are consistent with others who have found that the majority of low-income women feel that they have at least some control over condom use with their partner (Cabral et al., 1998; Soler et al., 2000). It is interesting to note that in 3 of the 5 events where condoms were available but not used, women appeared not to bring up the issue of using a condom because they had concerns regarding how it might influence the relationship with their partner. Our findings also seem in line with those of Weeks et al. (1995) who concluded that the scarcity of available men, combined with women’s reliance on these men for support, may tip the balance of power toward the male partner and erode impoverished women’s ability to engage in HIV self-protective behavior such as condom use. Because heterosexual intercourse accounts for 83 percent of women’s HIV/AIDS diagnoses (CDC, 2007), continued attention to condom use as an HIV protective device is necessary.

Interventions that hope to assist women in practicing safer sex may want to address women’s understanding of “trust” and “safety” with a partner as well as tackle ways to interact with men whose sexual behaviors are beyond their knowledge and control. It is clear from our diverse set of sexual episodes that there is a need to increase women’s recognition of partner risk and their ability to negotiate safer sex in ways that do not threaten the relationship. Given the diversity of circumstances that homeless women experience, interventions should target women in committed relationships as well as women who have sex outside of committed relationships. Such interventions should attend, for example, to commitment and its influence on condom use. Programs that emphasize communication and negotiation skills, strategies to develop assertiveness, and effective conflict resolution may be effective (e.g., El-Bassel et al., 2005; Theall et al., 2003).


This study faces several limitations. Although both shelters and women were selected randomly, the sampling design was such that population prevalence estimates could not be generated for homeless women who use shelters in Los Angeles. To be fair, however, our intent was neither to generate precise and accurate population estimates nor to generate formal models to predict safe sex practices. Instead, the goals of this exploratory research were to: (1) capture the range of women’s sexual experiences (including women’s thoughts, feelings, behaviors, and the social and physical contexts in which sexual events are embedded); (2) identify – at a relatively crude level – what kinds of experiences were most and least prevalent; and (3) to generate hypotheses about mechanisms that are likely to lead to unsafe sexual practices, including the role of alcohol and drugs. The confidence in our findings comes from the degree to which the patterns appear across a diverse sample of women and sexual events rather than from a more statistically representative sample. We believe that the combination of a purposeful and diverse sampling strategy, a comprehensive conceptual framework to guide our interviews, and a semi-structured format that allowed women to describe their experiences in their own words was both an appropriate and cost-effective approach for obtaining our goals.

Second, the use of recall data to identify behavioral patterns poses a number of challenges. On the one hand, it is possible that hindsight and social desirability may have affected women"s responses and on the other, it is possible that women might have confused some of the details trying to remember past events. Our sense, however, is that women were very forthcoming with their stories. The fact that many women provided graphic sexual descriptions and explicit unprompted discussion of sex for resources and drugs makes clear that they had few problems talking about sensitive issues. We encouraged women to tell their own stories in their own words and, more often than not, prompted them for more details and a fuller understanding of their experience rather than asking them to give rationalizations for their behavior. Further, whenever interviewers saw inconsistencies, they asked women for clarification. This often resulted in not only a more in-depth understanding of the situation but also more insight into the kinds of rules and schemas that the women used.


This work was sponsored by a grant from the National Institute on Alcohol Abuse and Alcoholism (R01 AA015301). Many individuals and institutions contributed to this work. First, we wish to thank the 28 homeless women living in temporary shelters in Los Angeles County who permitted us to interview them and provided candid descriptions of what they thought and felt about their relationships and risky sexual behaviors. We thank the homeless shelters for their help and collaboration to make it possible to reach these homeless women. We thank our frontline interviewers Crystal Kollross and Dionne Barnes who took on the complicated task of conducting all of the semi-structured interviews and the RAND Survey Research Group for their hard work in finding and coordinating with the shelter sites. We would also like to thank Hilary Rhodes for all her coding efforts and Leigh Rohr for her assistance in preparing this report. We could not have done it without all of you.


  • Amaro H, Raj A. On the margin: Power and women"s HIV risk reduction strategies. Sex Roles. 2000;42:723–749.
  • Bailey SL, Camlin CS, Ennett ST. Substance use and risky sexual behavior among homeless and runaway youth. Journal of Adolescent Health. 1998;23:378–388. [PMC free article] [PubMed]
  • Bassuk EL, Weinreb LF, Buckner JC, Browne A, Salomon A, Bassuk SS. The characteristics and needs of sheltered homeless and low-income housed mothers. Jama. 1996;276:640–646. [PubMed]
  • Bernard HR. Research methods in anthropology: Qualitative and quantitative approaches. Lanham, MD: AltaMira Press; 2006.
  • Brennan P. Focus on psychometrics: The kappa statistic for establishing interrater reliability in the secondary analysis of qualitative clinical data. Research in Nursing & Health. 1992;15:153–158. [PubMed]
  • Cabral R, Pulley L, Artz L, Brill I, Macaluso M. Women at risk of HIV/STD: The importance of male partners as barriers to condom use. AIDS & behavior. 1998;2:75–85.
  • Castañeda D. The close relationship context and HIV/AIDS risk reduction among Mexican Americans. Sex Roles. 2000;42:551–580.
  • Center for Substance Abuse Treatment (CSAT) Practical approaches in the treatment of women who abuse alcohol and other drugs. Rockville, MD: SAMHSA; 1994.
  • Centers for Disease Control and Prevention. HIV/AIDS Surveillance Report. Rev ed. Vol. 17. Atlanta: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention; 2005.
  • Centers for Disease Control and Prevention (CDC) A glance at the HIV/AIDS epidemic. Atlanta, GA: CDC; 2007. Jun,
  • Chaisson RE, Bacchetti P, Osmond D, Brodie B, Sande MA, Moss AR. Cocaine use and HIV infection in intravenous drug users in San Francisco. Jama. 1989;261:561–565. [PubMed]
  • Cohen J. Weighted Kappa: Nominal scale agreement with provision for scaled disagreement or partial credit. Psychological Bulletin. 1968;70:213–220. [PubMed]
  • Cooper ML. Alcohol use and risky sexual behavior among college students and youth: evaluating the evidence. Journal of Studies on Alcohol Supplement. 2002;(14):101–117. [PubMed]
  • Crabtree BF, Miller WL. A template approach to text analysis: Developing and using codebooks. In: Crabtree BF, Miller WL, editors. Doing qualitative research. Newbury Park, CA: Sage Publications; 1992. pp. 93–109.
  • Davis KC, Hendershot CS, George WH, Norris J, Heiman JR. Alcohol’s effects on sexual decision making: An integration of alcohol myopia and individual differences. Journal of Studies on Alcohol and Drugs. 2007;68:843–851. [PubMed]
  • Dermen KH, Cooper ML. Inhibition conflict and alcohol expectancy as moderators of alcohol"s relationship to condom use. Experimental and Clinical Psychopharmacology. 2000;8:198–206. [PubMed]
  • Dermen KH, Cooper ML, Agocha VB. Sex-related alcohol expectancies as moderators of the relationship between alcohol use and risky sex in adolescents. Journal of Studies on Alcohol. 1998;59:71–77. [PubMed]
  • Edlin BR, Irwin KL, Faruque S, McCoy CB, Word C, Serrano Y, et al. Intersecting epidemics--crack cocaine use and HIV infection among inner-city young adults. Multicenter Crack Cocaine and HIV Infection Study Team. New England Journal of Medicine. 1994;331:1422–1427. [PubMed]
  • El-Bassel N, Witte SS, Gilbert L, Wu E, Chang M, Hill J, et al. Long-term effects of an HIV/STI sexual risk reduction intervention for heterosexual couples. AIDS & behavior. 2005;9:1–13. [PubMed]
  • Fisher J, Fisher W. Theoretical approaches to individual-level change in HIV risk behavior. In: Peterson J, DiClemente R, editors. Handbook of HIV prevention. New York: Kluwer/Plenum; 2000. pp. 3–56.
  • Fromme K, D'Amico EJ, Katz EC. Intoxicated sexual risk taking: an expectancy or cognitive impairment explanation? Journal of Studies on Alcohol. 1999;60:54–63. [PubMed]
  • Fullilove RE, Fullilove MT, Bowser BP, Gross SA. Risk of sexually transmitted disease among black adolescent crack users in Oakland and San Francisco, Calif. Jama. 1990;263:851–855. [PubMed]
  • George WH, Davis KC, Norris J, Heiman JR, Stoner SA, Schacht RL, Hendershot CS, Kajumulo KF. Indirect effects of acute alcohol intoxication on sexual risk-taking: The roles of subjective and physiological sexual arousal. Archives of Sexual behavior. 2008 April 23; e-publication ahead of print. [PMC free article] [PubMed]
  • George WH, Stoner SA, Norris J, Lopez PA, Lehman GL. Alcohol expectancies and sexuality: a self-fulfilling prophecy analysis of dyadic perceptions and behavior. Journal of Studies on Alcohol. 2000;61:168–176. [PubMed]
  • Gutiérrez L, Oh H, Gillmore M. Toward an understanding of (EM)Power(Ment) for HIV/AIDS prevention with adolescent women. Sex Roles. 2000;42:581–611.
  • Harvey SM, Beckman LJ, Gerend MA, Bird ST, Posner S, Huszti HC, Galavotti C. A conceptual model of women"s condom use intentions: Integrating intrapersonal and relationship factors. AIDS Care. 2006;18:698–709. [PubMed]
  • Heron M. Deaths: leading causes for 2004. National Vital Statistics Reports. 2007;56:1–95. [PubMed]
  • Leigh BC. Alcohol and condom use: a meta-analysis of event-level studies. Sexually Transmitted Diseases. 2002;29:476–482. [PubMed]
  • Leigh BC, Ames SL, Stacy AW. Alcohol, drugs, and condom use among drug offenders: An event-based analysis. Drug and Alcohol Dependence. 2008;93:38–42. [PMC free article] [PubMed]
  • Leigh BC, Stall R. Substance use and risky sexual behavior for exposure to HIV. Issues in methodology, interpretation, and prevention. American Psychologist. 1993;48:1035–1045. [PMC free article] [PubMed]
  • Leigh BC, Vanslyke JG, Hoppe MJ, Rainey DT, Morrison DM, Gillmore MR. Drinking and condom use: Results from an event-based daily diary. AIDS behavior. 2008;12:104–112. [PMC free article] [PubMed]
  • Lincoln YS, Guba EG. Naturalistic Inquiry. Beverly Hills: Sage Publications; 1985. Processing naturalistically obtained data; pp. 344–356.
  • Logan TK, Cole J, Leukefeld C. Women, sex, and HIV: social and contextual factors, meta-analysis of published interventions, and implications for practice and research. Psychological Bulletin. 2002;128:851–885. [PubMed]
  • Los Angeles Homeless Services Authority. 2005 Greater Los Angeles Homeless Count. 2006. Retrieved January 8, 2008, from
  • MacDonald TK, MacDonald G, Zanna MP, Fong GT. Alcohol, sexual arousal, and intentions to use condoms in young men: applying alcohol myopia theory to risky sexual behavior. Health Psychology. 2000;19:290–298. [PubMed]
  • MacQueen K, McLellan E, Kay K, Milstein B. Codebook development for team - based qualitative analysis. Cultural Anthropology Methods. 1998;10:31–36.
  • Marin BV, Tschann JM, Gomez CA, Gregorich S. Self-efficacy to use condoms in unmarried Latino adults. American Journal of Community Psychology. 1998;26:53–71. [PubMed]
  • McDonnell KA, Gielen AC, O'Campo P. Does HIV status make a difference in the experience of lifetime abuse? Descriptions of lifetime abuse and its context among low-income urban women. Journal of Urban Health. 2003;80:494–509. [PMC free article] [PubMed]
  • McKirnan DJ, Vanable PA, Ostrow DG, Hope B. Expectancies of sexual "escape" and sexual risk among drug and alcohol-involved gay and bisexual men. Journal of Substance Abuse. 2001;13:137–154. [PubMed]
  • Milburn N, D'Ercole A. Homeless women. Moving toward a comprehensive model. American Psychologist. 1991;46:1161–1169. [PubMed]
  • Miles MB, Huberman AM. Qualitative Data Analysis: An expanded sourcebook. 2nd ed. Thousand Oaks, CA: Sage Publications; 1994.
  • Misovich SJ, Fisher JD, Fisher WA. Close relationships and elevated HIV risk behavior: Evidence and possible underlying psychological processes. General Psychology Review. 1997;1:72–107.
  • Morrill AC, Ickovics JR, Golubchikov VV, Beren SE, Rodin J. Safer sex: social and psychological predictors of behavioral maintenance and change among heterosexual women. Journal of Consulting and Clinical Psychology. 1996;64:819–828. [PubMed]
  • Morrison DM, Gillmore MR, Hoppe MJ, Gaylord J, Leigh BC, Rainey D. Adolescent drinking and sex: Findings from a daily diary study. Perspectives on Sexual and Reproductive Health. 2003;35:162–168. [PubMed]
  • Morse JM. Designing qualitative research. In: Denzin NK, Lincoln YS, editors. Handbook of qualitative research. Thousand Oaks, CA: Sage Publications; 1994. pp. 220–235.
  • Muhr T. Atlas/ti-A prototype for the support of text interpretation. Qualitative Sociology. 1991;14:349–371.
  • National Institute on Alcohol Abuse Alcoholism. Alcohol and HIV. Alcohol Alert. 2002;57:1–4.
  • Nyamathi A. Health Promotion among Homeless Women at Risk of AIDS. Rockville, MD: National Institute of Mental Health; 1994.
  • Nyamathi AM, Stein JA, Swanson JM. Personal, cognitive, behavioral, and demographic predictors of HIV testing and STDs in homeless women. Journal of Behavioral Medicine. 2000;23:123–147. [PubMed]
  • Padgett DK, Struening EL. Victimization and traumatic injuries among the homeless: associations with alcohol, drug, and mental problems. American Journal of Orthopsychiatry. 1992;62:525–534. [PubMed]
  • Patton MQ. Qualitative evaluation and research methods. 2nd ed. Newbury Park, CA: Sage Publications, Inc; 1990.
  • Pulerwitz J, Gortmaker SL, DeJong W. Measuring sexual relationship power in HIV/STD research. Sex Roles. 2000;42:637–660.
  • Ryan G, Bernard HR. Techniques to identify themes. Field Methods. 2003;15:85–109.
  • Schilling R, el-Bassel N, Ivanoff A, Gilbert L, Su KH, Safyer SM. Sexual risk behavior of incarcerated, drug-using women, 1992. Public Health Reports. 1994;109:539–547. [PMC free article] [PubMed]
  • Seal DW, Ehrhardt AA. HIV-prevention-related sexual health promotion for heterosexual men in the United States: pitfalls and recommendations. Archives of Sexual behavior. 2004;33:211–222. [PubMed]
  • Sikkema KJ, Heckman TG, Kelly JA, Anderson ES, Winett RA, Solomon LJ, et al. HIV risk behaviors among women living in low-income, inner-city housing developments. American Journal of Public Health. 1996;86:1123–1128. [PubMed]
  • Soler H, Quadagno D, Sly DF, Riehman KS, Eberstein IW, Harrison DF. Relationship dynamics, ethnicity and condom use among low-income women. Family Planning Perspectives. 2000;32:82–88. 101. [PubMed]
  • Spradley JP. The ethnographic interview. New York: Holt, Rinehart and Winston; 1979.
  • St. Lawrence JS, Eldridge GD, Reitman D, Little CE, Shelby MC, Brasfield TL. Factors influencing condom use among African American women: Implications for risk reduction interventions. American Journal of Community Psychology. 1998;26:7–28. [PubMed]
  • Steele CM, Josephs RA. Alcohol myopia: Its prized and dangerous effects. American Psychologist. 1990;45:921–933. [PubMed]
  • Testa M, Collins RL. Alcohol and risky sexual behavior: Event-based analyses among a sample of high-risk women. Psychology of Addictive Behaviors. 1997;11:190–201.
  • Theall KP, Sterk CE, Elifson KW. Male condom use by type of relationship following an HIV intervention among women who use illegal drugs. Journal of Drug Issues. 2003;33:1–28.
  • Tortu S, McMahon J, Hamid R, Neaigus A. Drug-using women's sexual risk: An event analysis. AIDS and behavior. 2000;4:329–340.
  • Tucker JS, Elliott MN, Wenzel SL, Hambarsoomian K. Relationship commitment and its implications for unprotected sex among impoverished women living in shelters and low-income housing in Los Angeles County. Health Psychology. 2007;26:644–649. [PubMed]
  • Vanable PA, McKirnan DJ, Buchbinder SP, Bartholow BN, Douglas JM, Jr, Judson FN, et al. Alcohol Use and High-Risk Sexual Behavior Among Men Who Have Sex With Men: The Effects of Consumption Level and Partner Type. Health Psychology. 2004;23:525–532. [PubMed]
  • Weeks MR, Schensul JJ, Williams SS, Singer M, Grier M. AIDS prevention for African American and Latina women: Building culturally and gender appropriate intervention. AIDS Education and Prevention. 1995;7:251–263. [PubMed]
  • Wenzel SL, Tucker JS, Elliott MN, Hambarsoomians K. Sexual risk among impoverished women: understanding the role of housing status. AIDS & behavior. 2007;11(6 Suppl):9–20. [PubMed]
  • Wenzel SL, Tucker JS, Elliott MN, Hambarsoomians K, Perlman J, Becker K, et al. Prevalence and co-occurrence of violence, substance use and disorder, and HIV risk behavior: a comparison of sheltered and low-income housed women in Los Angeles County. Preventive Medicine. 2004;39:617–624. [PubMed]
  • Willms DG, Best JA, Taylor DW, Raymond GJ, Wilson DMC, Lindsay EA, et al. A systematic approach for using qualitative methods in primary prevention research. Medical Anthropology Quarterly. 1990;4:391–409.
  • Wingood GM, DiClemente RJ. Partner influences and gender-related factors associated with noncondom use among young adult African American women. American Journal of Community Psychology. 1998;26:29–51. [PubMed]
  • Wingood GM, Scd, DiClemente RJ. Application of the theory of gender and power to examine HIV-related exposures, risk factors, and effective interventions for women. Health Education and Behavior. 2000;27:539–565. [PubMed]
  • Wood D, Valdez RB, Hayashi T, Shen A. Homeless and housed families in Los Angeles: a study comparing demographic, economic, and family function characteristics. American Journal of Public Health. 1990;80:1049–1052. [PubMed]