Methods. Data collection involved life histories and participant observation with migrant women in Atlanta and their sisters or sisters-in-law in Mexico.

Results. Both younger and older women acknowledged that migrant men’s sexual behavior may expose them to HIV and other sexually transmitted diseases.Younger Mexican women in both communities expressed a marital ideal characterized by mutual intimacy, communication, joint decisionmaking, and sexual pleasure, but not by willingness to use condoms as an HIV prevention strategy.

Conclusions. Migrant Mexican women’s commitment to an illusion of fidelity will hinder HIV prevention initiatives targeted toward them. Furthermore, the changing meanings of marital sex may make it harder to convince young couples to use condoms as an HIV prevention strategy. If the chain of heterosexual marital HIV transmission is to be interrupted in this community, prevention programs must target men. (Am J Public Health. 2002;92:1227–1237)

This article describes the social context of the migration-related HIV epidemic in western Mexico. In recent years, epidemiological and ethnographic research has suggested that an increasing proportion of AIDS cases in Mexico are related to Mexican men’s becoming infected in the United States and then returning home to infect their wives.2 A growing body of work has demonstrated cross-cultural variation in people’s ideas and emotions about desire, partner choice, and sexual behaviors,3 and in this article we apply these insights about the social construction of sexuality to study how social context shapes women’s risk of migrationrelated HIV and sexually transmitted disease (STD) infection. Specifically, we show how a shift toward a more companionate marital ideal among women from 2 migrant-sending communities in Mexico has created generational differences in women’s attitudes toward men’s sexual infidelity, and we argue that a growing emphasis on sexual intimacy within marriage may cause young Mexican women to be particularly invested in ignoring evidence of infidelity.

Previous studies of AIDS risk denial have demonstrated how women’s commitment to the illusion of fidelity can increase the risk of infection with HIV.4 Here, we document AIDS risk denial in a context in which it has not previously been noted, and we show how it may potentially be exacerbated by changing notions about marital intimacy. We also suggest that the same ideas that hinder women’s recognition of their risk may contain possibilities for culturally relevant prevention messages targeted at men.

As we trace the contours of sexuality in these migrant-sending communities, we discuss differences between older and younger women in terms of how they talk about sexuality, marriage, and infidelity. These generational differences call attention both to the importance of studying how sexuality changes across time and to the ironic possibility that increased communication between married couples about some aspects of sexuality (such as mutual pleasure) does not necessarily imply a greater likelihood of communicating about the risk of infection and the need for prevention. We also describe differences among women in how they interpret this discourse about marital intimacy and the meaning of infidelity, pointing out that changes in the social constructions of sexuality are not uniform across society. In our conclusion, we critique programs that aim to prevent heterosexual HIV transmission by working with married women and make some concrete suggestions on how these findings about sexuality and marriage could be used to shape a culturally relevant prevention campaign for Mexican migrants.

The data come from a comparative ethnographic study of gender, sexuality, and reproductive health among Mexicans from a migrant-exporting community in rural western Mexico.5 The broader study explored the effect of generational and migration-related change on gender and sexuality and the way these changes in gender and sexuality shape reproductive health practices. The primary method of data collection was life history interviews, conducted with 13 women in Atlanta and 13 in the Mexican communities.6 The life history informants in Atlanta were recruited from an initial set of research contacts from El Fuerte (a rural community in Michoacan state) and Degollado (a town in Jalisco state) who lived in the Atlanta metropolitan area. The life history informants in the Mexican field sites were the sisters or sisters-in-law of the informants in Atlanta. Tables 1 and 2 summarize the sampling strategy and ethnographic methods. The Mexican field sites consisted of Degollado and El Fuerte. The Census Bureau estimates that the Latino population of the Atlanta metropolitan area is over 175 000; other estimates place the number closer to 250 000.9

TABLE 1 —Summary of Sample Distribution for Key Informants (Life History Informants [LHIs] and Selected Mothers and Husbands of LHIs)

TABLE 2 —Summary of Ethnographic Methods Employed, by Population, Topics, and Sample Size

The data for this study consist of the transcribed interviews and field notes. The analysis involved multiple readings of the text. In the first reading, we generated topical themes (i.e., infidelity) and sorted sections of narrative into separate topical files for each code. In the second reading, we looked at the variability within topics (i.e., differences in what women said about infidelity) to establish subcodes. A third level of analysis involved exploring differences between the subcodes—that is, looking at the social differences between women who said they would never confront their husband if faced with evidence of infidelity and those who said that they might. For a more detailed discussion of analytic methods, see Hirsch and also Miles and Huberman.10

There is a long history of labor migration to the United States from western Mexico; some of those interviewed for this study had fathers and grandfathers who had traveled to the United States as temporary labor migrants.11 The proportion of Mexican migrants to the United States that is female has increased in the past decade,12 but significant numbers of married and unmarried men still travel north for 9 or 10 months every year from towns throughout Mexico, leaving their wives or girlfriends behind. As indicated in Table 1 , 14 of the 26 life history informants had experienced these extended marital separations.

Mexican migrants’ relative youth, their loneliness and social isolation, and the fact that many are single or traveling without their spouses may make them likely to seek sexual activity, and low levels of education and limited English may make it harder to reach these migrants with messages about HIV prevention.13 Social factors such as more permissive sexual norms, the higher incidence of HIV/AIDS (especially in the states that have been the destinations for great numbers of Mexican migrants), more widespread use of injected drugs, the lack of Spanish-language prevention campaigns, and migrants’ limited access to health services can also increase their vulnerability to HIV.14 Migration itself, which is a product of the political and economic relationships between the United States and Mexico,15 may put Mexicans at risk, because men who experience extended separations from their primary partners may be more likely to engage in extramarital sex.16

Migration has emerged as a risk factor for AIDS in rural Mexico (Figure 1 ). In 1995, 25% of rural Mexican AIDS cases were among men who had been in the United States, while only 6% of people accounting for the urban cases reported travel to the United States.17 In Jalisco (where Degollado is located), one study found that half of the individuals reported to have AIDS had traveled to foreign countries,18 and in neighboring Michoacan, where El Fuerte is located, 39% of those with AIDS had traveled to the United States.19 Because women account for 20% of rural cases but only 14% of urban ones, it seems especially urgent to explore how life in rural Mexico places women at increased risk.20 Many have argued that seasonal labor migration to the United states is responsible for these urban–rural differences in sex ratios.21 If migration-related heterosexual transmission of HIV is to be prevented, it is important to know more about marital sexuality in rural Mexico.

FIGURE 1 —Percentage of Mexican AIDS cases classified as rural.

Although at the time this study was conducted, the Mexican National AIDS Council data25 had two recorded AIDS cases for the county in which Degollado is located, informants mentioned several cases of people living with AIDS in Degollado and El Fuerte and other instances in which men from the community had reportedly died of AIDS in the United States. Almost all of the life history informants recounted having heard about other women who had gotten an STD other than HIV from their husbands; many of these stories (with no prompting about connections to migration) were about men returning from stints in the north.26 All of the women knew that illnesses can be transmitted through genital contact during sex, but they all denied having had personal experience with sexually transmitted infections.27

Almost all of the life history informants had heard that condoms prevent the transmission of AIDS and other sexually transmitted diseases, but women’s ideas about the technical, sensory, and symbolic aspects of condoms make it unlikely that they would push for condom use. About half of the women had tried condoms as a method of pregnancy prevention, but none favored them, and only a few used them occasionally in combination with rhythm.28 One reason for preferring other methods, these women said, is their belief that condoms are ineffective. Both those who had used them and those who had not made comments such as “no es tan seguro” (it’s not so safe or effective) and “a veces salen medio defectuosos” (sometimes they are defective). Women also talked about how condoms are a barrier to intimacy.29 One woman who had tried using condoms said, “I told him, to have sex every now and then and to do it with a cover, all covered up, no way.” Women spoke of not wanting to be “tapado” (covered up), and said that sex with a condom is less natural and less pleasurable (for both women and men). Women saw unprotected sex as the most intimate kind of sex, and thus argued that if a man was going to use a condom, he should use it with his other partners. One migrant’s wife said she had tried to get her husband to use condoms as a method of pregnancy prevention during a surprise homecoming, only to be told by him, “No, then what the hell kind of pleasure would that be? In that case it would be better to go with another damn woman.” Men and women reunited after a migrant separation are jokingly said to be de luna de miel, on honeymoon; the romance of this honeymoon period would be considerably dampened by the implication that condoms were necessary because either partner had strayed.

Given the risk posed by a strategy of unilateral monogamy, one approach might be to educate these women about their HIV risk and develop interventions to encourage them to negotiate with their husbands for condom use. Indeed, many of the articles cited in Gayet, Bronfman, and Magis’ review article on HIV and migration30 suggest that programs should focus on increasing women’s condom negotiation skills. In the remainder of this article, we argue against this seemingly obvious response. It is not just that women do not like condoms (although they may not) or that they may not have the power to insist on their use (although, again, they may not). Rather, negotiating for condom use is unthinkable to these women, because they see requesting condom use as tantamount to acknowledging or even giving permission for a husband’s infidelity.

Within the new ideal of the Mexican companionate marriage, young couples spend more time together, enjoy a somewhat less strongly gendered division of labor, and claim joint decisionmaking (although frequently with ultimate masculine authority) in many areas of domestic and economic issues. These companionate marriages are not necessarily egalitarian; in fact, the emphasis on the importance of marital ties above all other social connections may make women more isolated and dependent on their husbands. Young women also speak differently than their mothers did about the nature and role of sexuality within marriage. For the older women, marital sex produced children and held a man’s attention, and thus served to direct his resources toward a woman and her children rather than outside the family. The younger women, in contrast, claim that mutually pleasurable sexual intimacy strengthens the confianza that is at the heart of these modern marriages.32 They see sex as a sort of “marital glue,” helping a couple’s affective relationship survive through good times and bad.33

The younger Mexican women in Atlanta shared values about relationships, sexual intimacy, and gender with women in the sending communities, but women in the Atlanta field sites live in very different social circumstances than do their sisters and sisters-in-law in Mexico, and so some women in Atlanta had a great deal more domestic power than did their sisters or sisters-in-law in Mexico.34 The contrast between the shared culture and the distinct social settings of this type of migrant community provides a unique opportunity to explore the way that social and cultural factors shape sexuality.

Younger women, in contrast, see men’s sexual behavior as inseparable from marital intimacy. If sex is the language of love, then to be sure of their husband’s love they must believe that he is faithful while away. As one young woman in the rancho said, “I wouldn’t feel comfortable with him, knowing he’d been going around with another woman.” A man’s sexual infidelity, these younger women said, would weaken or destroy the confianza they have with their husbands. One younger woman told a story about calling her husband in Atlanta from Mexico. The phone rang all night, but he never answered. When she accused him the next day of having been out with another woman, he said he had been home and just not heard the phone. She reflected on this incident with some general comments about infidelity:

Women’s commitment to the image of their husbands as sexually faithful, even when confronted by compelling evidence to the contrary, draws on the idea that sexual betrayal indicates a lack of love.37 The new cultural ideal of marital companionship and shared sexual pleasure gives younger women yet another reason not to raise issues of infection and infidelity. The younger couples prided themselves on their communication about sexuality, emotions, and pleasure, in contrast to the older women, for whom open discussions about sexual matters indicated a lack of respect. However, the value these younger couples place on communication about some aspects of sexuality does not mean that the younger women are any more likely than their mothers were to suggest that a man use a condom for disease prevention purposes.

Six of the 26 life history informants said that infidelity was grounds for leaving one’s husband.38 Only the women who could support themselves adopted this rights-based language to talk about infidelity39; they all had an independent income, and 5 had at least a high school diploma as well. When asked about men who said that extramarital sex was their right, one woman who lived in Atlanta and contributed as much as her husband to their joint budget said,

She argued that in the emotional contract of marriage, both partners have essentially the same set of rights. When asked why women do not put up with infidelity now, she continued, “Before, women did what their husbands said when they got married. The rule was that here the man was boss. . . . Now we both are the boss.” These Mexican women have woven together ideas about sexual intimacy as the foundation of marriage with their personal histories and their self-image as modern women to produce a distinct set of attitudes about where to set the limits on men’s sexual freedom.

Others said that they might address the issue by trying to discuss it with their husbands. (The “talk it out” group presented a strategy for dealing with infidelity when forced to do so, not for raising the issue in a preventive health context.) Talking about how she would counsel a friend to act, one said, “The thing is to make him understand, communicate well with her, so that he realizes that his wife is worthwhile.” These women share with their sisters the idea that infidelity is a fundamental violation of the marital bargain, but these women seemed particularly skilled at talking about their feelings, and they prized the open communication they have achieved in their marriages. Moreover, they all resided in Mexico. The fact that women in Atlanta and wealthier women in Mexico are so forthright in saying that they would leave their husbands, while their matches in Mexico say that they would do their best to talk things through, shows how emotional closeness can be a strategic resource for a woman who is socially and economically dependent on her husband. Women in Degollado or El Fuerte may have some resources that their peers in Atlanta lack (e.g., broader social networks, a mastery of the language), but these resources do not enable women to live without a man’s economic support. Furthermore, women in Mexico continue to face the issue that “el hombre es el respeto de la casa,” that a woman needs a man to be respected,40 as well as the stigma of divorce, still strong in rural Mexico.

The idea that these women in Mexico would try to talk it out implies ultimately accepting a man’s infidelity (just as their mothers might have), but saying that they would at least confront the issue allows them to assert that they are women who deal actively with problems and who are free to speak their minds. For both groups of women, infidelity represents a betrayal of the special sexual relationship that married people should share, but the 2 groups use the ideology differently, based on their social and economic circumstances.

Not all women, of course, were so thoroughly invested in these modern strategies for building relationships. The idea that a successful relationship depends on a woman’s ability to keep her man happy, to work through his good side (manejarlo por las buenas), was alive and well among Mexican women in both US and Mexican field sites. The 4 women in this third group focused much more on their feminine ability to please and hold onto a man. A satisfying sex life, they all said, was one of the keys to a good marriage. One responded to the question “If a couple does not get along well in bed, how will they do in the rest of their marriage?” by saying, “Badly, I think, because I have heard that for the man [you need to] have his food ready, his clothes ready, and the bed ready, and he will be all content, and if not, then [he’ll be] a real devil.” In the event of actual or suspected infidelity, these women said that they would use their feminine wiles to make themselves irresistible. All of them mention the importance of being more cariñosa, sweeter or more loving. The women in this group are young (under 20) and recently married. Although they lived in both the US and Mexican field sites, none worked outside the home, and the ones in Atlanta were there without any of their own blood relations to count on in an emergency. Their support for this strategy is most likely a product of many factors: age, family histories, their personality and their husband’s, the sort of marriage they have, and their lack of other resources.

Finally, 6 of the life history informants said that they would do nothing in reaction to suspected or confirmed infidelity. None relished the thought, but neither did she argue that it would be grounds for separation. Although this group included women interviewed both in Atlanta and the Mexican field sites, those in Atlanta had spent some, if not all, of the years of their marriage in Mexico; all of the women in this group were older, less educated, with less work experience and fewer opportunities for economic independence. In other words, the women whose material circumstances were most similar to women of the older generation were least likely to embrace this new ideology of sexuality.

The younger generation’s new ideas about sexual intimacy transform men’s infidelity from a painful reminder of gender inequity—but not necessarily an indication that a marriage is a failure—to a betrayal of the confianza that young women hope lies at the core of their marriage. If infidelity represents a breach of trust—and an STD or HIV is the ultimate evidence of that breach—then every act of sex without a condom is a mutual performance of trust. The confianza of which younger men and women speak implies a shared commitment to pleasure and emotional intimacy, predicated on the unquestioned and unquestionable assumption of mutual monogamy. It does not mean that they are open to discussing issues of risk and infidelity or negotiating condom use. In a marriage with plena confianza (complete trust), a woman might feel freer to initiate sex, to have oral sex, or to don some racy lingerie, but she is no more likely than her mother was to ask spontaneously about her husband’s adventures in el norte.

Whether in Atlanta or in the Mexican field sites, younger women expressed this commitment to marriages of confianza; younger women in Mexico were more similar to younger women in Atlanta than they were to older women in Mexico. The greater a woman’s social and economic resources, however, the less likely she was to say that she would try to work it out if confronted with hard evidence of infidelity. Women’s imagined responses show that social stratification has an important effect on the options women see as being available to them (Table 3 ).

TABLE 3 —Main Findings of This Study, by Topic

Developing effective strategies for HIV and STD prevention among married women whose husbands are seasonal migrants will mean finding a way to work around women’s HIV risk denial. Asking a man to use a condom implies mistrust and, as one young woman in Atlanta (whom gossip suggests has good reason to worry about her own risk of contracting an STD) said, “Mistrust is a lack of love” (“La desconfianza ya es falta de amor”). Sobo41 found that African American women in the United States held similar views, and she argued that women’s commitment to the ideal of monogamy contributes to their risk of HIV infection.42

Our data here support Sobo’s conclusion, but we also extend her work by placing married women’s AIDS risk denial within the context of broader global changes in the meaning of marriage. Ethnographic research from across the developing world has described the evolution of similar ideals about marital companionship and sexual intimacy.43 The similarity of these ideals does not mean that marriage in El Fuerte is exactly the same as marriage in Egypt, but the widespread link between marital sex and emotional intimacy suggests that married women’s AIDS risk denial, combined with persistent gender inequality, may have greater implications for the spread of marital HIV transmission than has previously been acknowledged.

The continuing focus, however, on why women will or will not negotiate for condom use represents an astonishing failure on the part of public health researchers to confront how gender shapes—and constrains—our imagination. The focus on women may be in part a product of scientific knowledge about women’s greater biological risk of sexual infection with HIV,44 but this does not explain why we have assumed that women can modify their sexual behavior to press for condom use but that men cannot,45 nor does it justify the fact that we have continued to push the message that monogamy is an effective HIV risk reduction strategy46 when we know that the assumption that marriage equals monogamy may be costing women their lives.47 Men’s sexual behavior presents a public health problem, and thus these men’s behavior is the proper object of our attention. To suggest that we can help married women protect themselves by “empowering” them to negotiate for condom use is to suggest that we can change the outcome of gendered inequalities in power without doing anything about the actual inequality.48

Given all that a woman in this community might feel that she stands to gain by ignoring her risk of contracting an STD from her husband, it is small wonder that women do so little to protect themselves. These women would not benefit from programs to help them with condom negotiation; in many areas of their domestic and wider social lives, they are already expert negotiators. Women in this community do not want to use condoms for disease prevention because the emotional cost of acknowledging that sex with their husbands is not safe sex is simply too high. As Karen Mason has pointed out, the idea that women’s empowerment would increase condom use “is plausible only if there is reason to think that women are more motivated than men to use condoms, but are unable to enforce their use in the face of male opposition.”49 In this particular Mexican case, the cultural logic of marital sexuality suggests that public health efforts to stem a rising tide of migration-related heterosexual transmission of HIV in Mexico should instead concentrate on 2 fronts: the behavior and social circumstances of migrants while they are in the United States, and the wider sources of gender inequality that structure women’s lives.50

The research presented here could serve as the basis for a prevention program that builds on the ideas of respect and trust (respeto and confianza) to shape health education messages targeted toward men to promote extramarital condom use. Men and women may not agree on whether men who profess to believe in companionate marriage should ever have outside partners, but they do agree that for a woman to find out about her husband’s infidelity represents both a betrayal of trust and a lack of respect on his part. Health education messages could use these terms to emphasize a man’s responsibility to protect his family from disease and hurt feelings. Telling men that ideally they should have no outside partners, but that if they do they should use a condom to protect the confianza they have with a spouse and to avoid showing her a falta de respeto, a lack of respect, may be one way to frame this important health education message in culturally meaningful terms.51

Our call for HIV prevention programs for Mexican migrants needs to be considered in the context of Mexican migrants’ generally limited access to health services. In Atlanta, the only way a Mexican migrant is likely to get attention from the Fulton County Health Department is if he has active tuberculosis.52 Nationally, Latino immigrants’ limited access to quality US health care services has been well documented.53 At the time of the writing of this article, most of the men who were laying bricks and hanging Sheetrock in the ever-larger suburban minimansions that ring Atlanta, along with their paisanos mowing lawns and planting seasonal flowers, had very limited access to primary health care. These men who spend long months in the cold north may be powerful in relation to their wives, but we should remember that they are not just men—they are poor, dark-skinned, primarily non–English-speaking, sometimes undocumented men. Addressing migration-related HIV in anything other than a superficial way will mean facilitating access to health care in a more general fashion. Affordable clinics with translators and evening hours could create a wide range of opportunities for primary and secondary prevention of STDs and for HIV prevention.54

AIDS is an emergency, and the sort of research presented here takes a long time (although partnering with local social scientists and beginning research projects with a firm grounding in the vast ethnographic record can certainly speed this up). However, it is precisely because of the pressing nature of the global pandemic that we must take the time to do careful ethnographic research. In particular, the data presented here speak to the value of anthropological research for drawing connections between culture and changing patterns of social inequality. As we enter the third decade of the epidemic, we must use all the tools within our grasp, including the full range of anthropology’s explanatory power, to address the global challenge of AIDS.

J. S. Hirsch gratefully acknowledges support from the Andrew Mellon Foundation through a grant to the Department of Population Dynamics at Johns Hopkins University; the National Science Foundation Program in Cultural Anthropology (SBR-9510069); the International Migration Program at the Social Science Research Council; the Emory AIDS International Training and Research Program, which is funded by a grant from the Fogarty International Center at the National Institutes of Health (1 D43 TW01042-02); and the Department of International Health at the Rollins School of Public Health. J. Higgins gratefully acknowledges financial support from the program in Women’s Studies at Emory University.

We thank Claire Sterk and Carlos del Río and 3 anonymous Journal reviewers for comments and Carlos Magis Rodríguez of the Consejo Nacional de SIDA for generously sharing surveillance data on rural AIDS in Mexico. Our deepest debt is to the women and men who opened their homes and their hearts throughout the course of this study; we cannot acknowledge them by name, but we hope at least to have been true to their stories.

Human Participant Protection
The study described here received approval from the institutional review board at Johns Hopkins School of Hygiene and Public Health, and all subjects provided informed consent before participating in interviews.