Version classiqueVersion mobile

Sexualités américaines

Claudine Raynaud

Sex in America: Politics, Science, and Response to a National Survey

Stuart Michaels

Texte intégral


1Over 7 years ago, in January 1988,I began working with a group of colleagues at the University of Chicago on a project to develop a survey of sexual behavior in the United States using modern state-of-the-art survey research methods. The results of that survey were finally published in October 1994 in two books, one academic, The Social Organization of Sexuality (Laumann et al., 1994) and one for a more popular audience, Sex in America (Michael et al., 1994). This caused a minor media storm until it was pushed from the front pages by the beginning of the O. J. Simpson trial and in intellectual circles by the publication of The Bell Curve and then by the Republican sweep of the congressional elections in November. I would like to reflect on this process in the hopes of shedding some light on issues related to sex and gender in the United States.

2There are many different ways to think about this survey. One of the apparently simplest and most naive ways to view it is as a direct reflection of the sexual behavior of adults in contemporary American society. One could say that is how the media treated the study, though they tended to view it as a kind of empirical report-card or moral barometer, rather than as a more neutral comprehensive quantitative investigation of the variation and distribution of various aspects of sexuality across the population. Both in order to understand this study and to understand sex and gender in the U. S. today, we would do much better to take a step back from the specifics of the survey findings themselves and look at them in the context — historical, intellectual, and political — of how and why the survey was done. That is what I would like to try to do today: to present some of the background of the study, particularly its checkered political history, then to talk a little about the methods and findings of the survey itself primarily to contrast them with the media treatment of the study. Along the way I will try to make some preliminary comparisons with the similar studies done in Britain and France which were completed before the American research.

The Politics of Sexual Behavior Research in the U. S. in the Age of AIDS

3To understand this research project — why it began when it did, why it took the form that it did, the political attacks that it suffered that finally ended the governmental funding — one needs to look at its immediate history.

  • 1 This process is described in a series of publications from the National Academy of Sciences in Was (...)

4This background is not as simple and straight-forward as one might imagine, especially given the subsequent history and how politicized sex and sexual issues have become in the U. S. In particular, it is ironic that this research originated during the Reagan administration and initially was actively being pushed by the executive branch. The American government began to seriously push for a major study of sexual behavior as part of the response to the AIDS epidemic in 1987. The outlines of that response had been being drawn up by blue-ribbon scientific panels organized in 1985.1 In this period government spending on research issues related to HIV/AIDS and treatment and care of people with HIV/AIDS expanded rapidly to close to their current levels.

  • 2 This process with its peculiarly American emphasis on voluntarism and self organization, in the fa (...)

5Six or seven years after the identification of the syndrome, the conservative administration and the congress were finally responding to the crisis. From a stance of almost total neglect, the government was quite suddenly in a relatively active, aggressive mode. By this time, it was clearly realized that a medical cure for AIDS (or prevention through a vaccine) would be a long time coming and that instead prevention through behavior change would have to be the primary public health response. There were advances in treatment of persons already infected, but radical measures were also needed to help prevent the further spread of the epidemic. This would depend on education and behavior change. (I should point out that this had already begun and had largely been accomplished within the gay communities, the epicenters of the epidemic in the United States.)2 But many questions remain about the likely magnitude and direction of the epidemic, questions which depended on knowledge about how large the population that was connected to the already existing epicenters of the epidemic (largely made up of gay men in major cities and intravenous drug users) might be. How many men had sex with men but were not found in the studies that had been organized around the clienteles of gay clinics and organizations? How large was the so-called "bisexual bridge," i. e., men who had sex with men in populations with high levels of HIV and also had sex with women? What was the distribution of numbers of sexual partners and sexual practices throughout the population? Basic information about sexual behavior, both for policy, e. g., for health resource allocation and for identifying sites for prevention and education, was sorely lacking. Shockingly, the only serious scientific effort to measure and describe the distribution of sexual practices in the American population was the Kinsey research that was over 40 years old. And furthermore Kinsey's data had not been based on a statistically representative sample. A key component of a scientific research program would obviously be to rectify this glaring lack of basic information.

6Similar national research projects were begun at almost the same time in a number of European countries: the Netherlands, England, France, Norway, and Finland. While in one sense, AIDS focused all the research efforts on certain aspects of sex — penetrative sex practices, condom use, homosexuality, numbers of partners — each team brought its own intellectual and disciplinary biases to the research. Perhaps a very quick sketch might help illustrate this. The English project was led by epidemiologists and public health professionals including several with an emphasis on health education. This led to a rather narrowly focused health and risk practice approach and one that was very behavioral. The Dutch team, in contrast, was largely drawn from sexology. This led to a greater emphasis on early sexual experiences, some broader interest in practices such as masturbation and comfort with sexuality. The French team was the most eclectic including epidemiologists (its leadership from INSERM) but including sociologists and psychoanalysts as well. This led to both a high degree of focus on risk, but also interest in attitudes to death, love, and an interest in discussion networks.

  • 3 The epidemiological models of disease prior to AIDS were not rooted in the notion of networks, but (...)
  • 4 One of our key findings is that "homophily," the principle of "like attracts like" that shapes man (...)

7Our team was dominated by social scientists — sociologists and demographers. The leader and prime mover on the Chicago team is Edward Laumann, a sociologist and one of the leading proponents of social networks as a key organizing principle for the study of social life. A social networks perspective emphasizes the concrete relations between social actors (either individuals or larger scale social entities such as organizations). Unlike the sexological tradition which is either primarily psychologically oriented, closely tied as it is to clinical disciplines and practices (psychiatry, sex therapy) that conceive of and treat individuals as existing independently from any social context, or is more ecological or zoological as in the Kinsey tradition that viewed human sexuality as merely a subset of "mammalian" sexuality, the demographic and epidemiological questions related to the transmission of sexually transmitted diseases through the population practically cry out for a more social scientific approach, with a social network approach a particularly useful one.3 One of the major practical consequences of the social network approach in a survey of sexual behavior is to focus attention on the dyad rather than the individual actor as the key unit of analysis. While all AIDS related sex research has been concerned about partners to some extent, we focused to a much greater extent on specific partnerships. The enumeration and social description of sexual partners was a key component of our questionnaire. As much as possible we asked about sexual activities including issues such as contraception and prevention of disease within the context of specific relationships. This also led to a focus on the selection of sexual partners.4 Social networks include not only the network formed by sexual relationships, but the social networks, particularly friends and kin who surround and form the primary "audience" of interested parties to sexual and intimate relationships. I will return to some of these issues when I discuss the findings and results, but first let me turn to the more immediate political context.

8While by 1988 many agencies of the government were finally pushing for the collection of sexual behavior data, there were some hesitancies, including the unwillingness to allow official government interviewers associated with the bureau of the census to ask these sorts of questions of the American public. But the real resistance and reaction developed after the project was under way in 1989 and came from extreme right-wing politicians in congress, Jesse Helms being the loudest and most vociferous.

9It is not surprising, especially in retrospect (from attacks on funding for the arts, the 1992 Republican convention, anti-gay referendum campaigns in Oregon and Colorado and, most recently, the November 1994 congressional election that brought Newt Gingrich to the fore), that the effort to do a national sexual behavior survey became a target of the right wing in Congress and for the closely related largely church based grassroots. There are no doubt many reasons for this: the right's emphasis in its organizing on what it calls "family values" or traditional values, an emphasis on issues related to family, gender roles, opposition to abortion and especially resistance to the legitimation of homosexuality. The gay movement and gay rights were reemerging as a central target of the right, so it shouldn't surprise us that the combination of AIDS research and sex — especially the sexual behavior of all adults, not just gay men or prostitutes — would attract their attention.

10However, it is important to point out that the right-wing and especially the state is not monolithic. As I mentioned, it was under the Reagan administration that there was (eventually) a concerted push to collect basic information as part of the response to AIDS. During the initial phase of our project, the Office of Management and Budget (OMB), a part of the executive branch that has to approve all government documents, including research questionnaires, was asking why a major health survey that was planning to ask AIDS knowledge and attitudes questions did not have any sexual behavior questions, in particular, questions about anal intercourse. This survey did not add such questions because the census bureau (that not only collects the decennial census data but has a huge survey research field operation and collects data for a number of different agencies) would not allow its interviewers to ask these questions. Therefore, several government health agencies came to us and asked us to speed up our efforts so that we could collect the information for them. It was almost a year later, when we were ready to collect data, that the right wing became aware of our efforts — this was at the very beginning of the Bush administration. And their grass-roots lobbying, no doubt partially orchestrated from within the executive branch, focused on OMB. Thousands of letters poured into this obscure office of government. Right-wing newspapers and magazines and radio-shows ran stories denouncing us. Right-wing congress people sent letters to their colleagues and made speeches on the floor of the legislature.

11The right wing accused us (in advance) of having a political motivation. Their vision of this motivation was largely couched in homophobic terms. They feared that our primary goal was to advance the cause of homosexuality and they fantasized that we would do this by artificially increasing the percentage of the population that was gay. (Ironically, as it turns out, since one of the most consistent effects of studies that use modern survey research technologies, including carefully selected statistically representative samples, is to find lower rates of homosexuality — no matter how defined — than the Kinsey research which has served as the baseline for both scientific and popular thinking about these issues in the United States.) A long political battle ensued about whether or not the U. S. government would allow us to collect data — delaying the research by 2 to 3 years. This battle went on both within the bureaucracies under the Secretary of Health and Human Services and within Congress. There are many interesting stories to be told and lessons to be learned about the internal conflicts and processes of government and politics. I will not go into them here. Suffice it to say that while there was no question within the public health establishment about the need for such a study, no political force could stand up to the small extreme right in Congress that was able to define support for this research as support for pornography and homosexuality. These debates were very similar to what was happening at the same time in the arts arena with the attacks on the National Endowment for the Arts. They both involved a politicization of processes of peer review that had largely been insulated from direct political oversight. They also drew on populist rhetoric and strands of anti-elitism and anti-intellectualism widespread in American society — but most especially on evoking fears of threats to traditional values by "sexual radicals" mostly having to do with homosexuality.

12The final straw for government support of our research was in 1991 when a vote on future funding for a national survey was defeated by a vote of 66 to 34 and the political noose was tightened around the government scientists and administrators at the National Institute for Child Health and Human Development who had been the main funders of our work. These science administrators felt forced to suppress funding of a scientific peer reviewed grant proposal that had already been awarded, again much like the withdrawal of grants to the arts. At that point, we were able to secure funding from a set of private foundations in the U. S. that specialize in issues related to public health and were able to proceed again with the research though with a scaled down national survey. I should point out that our original design for a full-scale national survey called for an extremely efficient sample of 15,000 persons. The private foundation support allowed for a survey of only 3,400 persons about 1/5 the size. The national surveys in England and France were both on the scale we had suggested based on about 20,000 interviews in France and 18,000 in England.

13After waiting for about four years, and released from the internal politics and constraints of operating as agents of the government, we were almost relieved to be able to go forward with this much smaller, but less constrained survey. In the process of trying to construct a more acceptable survey for the government agencies we had been forced to delete questions about fantasy, masturbation, and any questions for couples that said they were married and monogamous for the past 10 years. The privately funded survey with the censored questions included was carried out over a 9 month period in 1992 and the final results appeared in October of 1994.

Results of the Sex Study

14What did we find? When we first began this research we did not really know what to expect. The Kinsey research published in the 1950s was basically all that existed in terms of comprehensive empirical findings that describe the sexual experiences of a non-clinical population intended to be broadly representative of the whole breadth of the U. S. population (Kinsey et al., 1948; Kinsey et al., 1953). Kinsey's findings had shocked American society in the 1950s. It was as if he had exposed a whole hidden sub-continent of sexuality that had never seen the light of day: masturbation, oral sex, homosexuality, and extra-marital sex were "discovered" to be extremely widespread. While a pioneering achievement in their day, by the mid 1980s, we knew Kinsey's methods were fundamentally flawed. His sample was not representative as he believed. But we did not know substantively what difference that made. Were his estimates too low or too high? What little bit of representative sample based research that had been done in the 1970s was primarily focused on women (usually younger women) and fertility. Thus, we knew that the age of first intercourse had been declining over time, but not much else.

15Since Kinsey, a number of very visible changes and events had taken place in the U. S., the "sexual revolution" of the 1960s and 1970s, the women's movement, the gay/lesbian movement, and an explosion of new kinds of talk about sex, much of it tied to the counter-culture of 1960s, and a mushrooming of various forms of therapy and self-help (Irvine, 1990). In the 1980s, AIDS generated a tremendous amount of attention to specific sexual practices, some of which, such as anal intercourse, had hardly ever been mentioned outside of very specialized circles. Thus, I think almost everyone expected that the results from a major new sex survey would be beyond anything that had been found 40 years earlier by Kinsey. No doubt this was a large part of the fear of the right wing, and perhaps, a hope of the sexual progressives for this kind of research.

16In this context and with these expectations, it is perhaps not surprising that the most striking impression that the ensemble of the results from this research made was of extremely moderate, even subdued rates of sexual activity. The numbers of partners that were reported, the frequency of sex itself, the incidence of masturbation, rates of same-gender sex, rates of extramarital sex were all much lower than any of us would have predicted.

17I had begun to get inklings of what we might find a few years before our survey was carried out and I must admit that my own initial reaction was one of disbelief. Looking at the results from a small set of questions on numbers of partners and the gender of these partners we were able to include on a related national survey done in 1988, I wondered if it might be a fluke (Davis and Smith, 1988). This survey found (and later surveys including our own have all found the same) that fully 80 percent of U. S. adult population reported at most 1 sexual partner in the last year (20 percent of these reporting no partners at all). Only 2. 5 percent of the men and about 1. 5 percent of the women reported a sex partner of their own sex in the last year. Even if these were somewhat under-reported by the persons interviewed, these numbers seemed too low, especially given the results of Kinsey's research that had been repeated verbatim and in summary form over and over by the developing gay rights movement. In particular, the idea that 10 percent of the population was homosexual or gay (exactly what was meant by this statement was never really specified) was widely accepted in the gay world and had been more or less accepted by the media. I thought that maybe the 1988 sample was skewed in some strange way. Maybe there was something about the language of the questionnaire. It took some time, including seeing the results from other studies in the U. S. and Europe, all of which used probability samples but in other ways varied the techniques and question wordings, until I was convinced that these results were not flukes. There was a very clear methodological consistency — technically one would say, these results were reliable. Using approximately the same technique they could be reproduced over and over again. But of course, the question remained, "Were they valid?" Did these results reasonably closely reflect the experiences of the population that was being studied and described? This is a complicated question that leads in many different directions. What I think is more interesting and useful here is to look at issues related to what was made of these findings, how they have been understood.

Media Response to the Sex Study

18How did the media interpret these and other "facts" or results from the study? If you saw the Time magazine "Sex in America" issue that covered the publication of the findings from our study in 1994, you will have a pretty good idea of most of the initial reaction to the study (Elmer-Dewitt, 1994). On the one hand there was a tremendous level of credulousness. The media accepted our word that this was the most comprehensive, most scientific, and most important sex study since Kinsey. Although I found this very satisfying, I also found it somewhat frightening. I think all of those statements are true (though perhaps complicated in their meanings). But I wondered how careful and skeptical the media had been in adopting, pretty much verbatim our own characterization of the research and broadcasting it to the public. What if we had not been a reputable group, what if our study had not been done extremely carefully, using advanced methods, and at great expense? Would any published account, authored by professors be taken equally seriously? I think segments of the media were themselves not completely certain — especially television reporters. On the one hand, there was the repetition of the importance and scientific standing of the study; on the other hand, news-readers would express their own, more personal skepticism about the research and in some ways dismiss it as yet another book or study with yet another interpretation, with absolutely no differentiation as to seriousness, quality, approach, etc. I remember, in particular, that there was an author who was making the news and talk show circuit at about the same time as our work was published. He was making some sort of sociobiological argument about the innate basis of the masculine drive to mate with as many partners as possible. Often, at the end of reports about our research, an almost spontaneous reference to that earlier research would be added with a tag line like, "Well who can you believe?" The media accounts demonstrate some of the highly stylized and simplistic ways that sexuality is treated. The sheer volume of attention confirms to what extent stories about sex are still titillating — in some very real sense, it is still true that "sex sells." Dress it up in slightly different clothes, promise a certain amount of scientific seriousness and a different point of view and even a heavy, scholarly, fairly technical approach can find itself on the covers of Time, U. S. News & World Report, Glamour (the latter two with sexy, soft-focus pictures of couples in bed) — featured on major television newsmagazines such as 20/20.

19Perhaps the most striking thing to me about the media response, was how the basic quantitative summary results about sexual behavior from our research, or generalizations about the relatively low rates we found compared to our own expectations, and what one might expect based on media discussions and portrayal (talk shows, made for TV movies, situation comedies, etc.), were turned into a moralizing account of the state of American society. The study became "good news." It was implicitly contrasted with some image of moral decay. This is clearly illustrated in some of the initial headlines and titles for news stories: "Sex study shatters kinky assumptions" (Chicago Tribune, October 6, 1994); "Sex in America: Faithfulness in Marriage Thrives After All" (New York Times, October 7, 1994); "Turns out we are sexually conventional" (USA Today, October 7, 1994); "Americans' Sex Lives Get Healthy Report" (October 7, 1994); "So, Now We Know What Americans Do in Bed. So?" (New York Times, Sunday Week in Review, October 9, 1994); and Sunday edition of the Chicago Tribune ran a second story with the title "What is normal? Sex in America. What the University of Chicago Survey Says" that had a sidebar entitled "Generation X is not Generation Sex" (October 9, 1994).

20What was much less likely to be discussed were findings which help to explain the relative low numbers of sexual partners — the fact that most sex takes place within relatively enduring relationships — a pattern that has long been recognized and is often referred to as "serial monogamy" — along with some of the attitude questions from the survey on satisfaction with given sexual relationships — was turned into a celebration of heterosexual marriage. A celebration that seemed to express a kind of surprise and relief. In most of the media accounts this was the dominant, organizing framing. The rest of the findings from the study were either used to fit into and support this view or they were largely ignored.

  • 5 We did not use the term heterosexual, instead we asked about the first time the respondent had vag (...)
  • 6 In addition, about 20 percent of the women say they had intercourse with a man for the first time (...)

21However, these views tended to obscure a number of equally impressive findings, especially some extremely disturbing findings that underscore the contrast between men and women's experience of sex. When we asked men and women about the circumstances of their first heterosexual experience, over 90% of men said that it was something they wanted to happen with almost all the rest saying that they went along with it even though they didn't necessarily want to do it at the time.5 However, only 70 percent of the women said that their first heterosexual experience was something that they wanted at the time, a quarter (25 percent) said they went along with the experience, and 4 percent said they were forced to have sex the first time. Even among the group who said they wanted to have sex the first time, the gender difference is marked: almost 50 percent (48%) of the women said they had sex the first time because they were in love compared to only 25 percent of the men. If the largest group of these women who actively desired their first intercourse said it was out of love, the larg est group of men (about half) said they did it because they were "curious/ready for sex", a response selected by only 25 percent of the women. Over 10 percent of the men say they did it for physical pleasure compared to 3 percent of the women.6

22In spite of the fact, that these sorts of statistics, underscoring the differences in men and women's experiences of sex, can be found throughout our book, they rarely played a role in media accounts and when they did they tended to be downplayed. Perhaps the most disturbing finding, and one that did receive some media attention, was the fact that 22 percent of the women reported ever having been forced to do something sexual by a man, whereas only 3 percent of the men report ever forcing a woman do something sexual she didn't want. One might think that this is due to a small number of men who rape many, many women. However, women classified only 4 percent of the men who forced them sexually as strangers (that is, 96 percent were known to the women who reported being forced; with 10 percent being husbands, 45 percent were men the women were in love with and 22 percent men the women knew well; only 20 percent were merely acquaintances).

23The major explanation appears to be that the men and women have very different impressions about what is occurring in these events. These results and many like them were almost totally ignored by the major media. If reported at all they were buried. They certainly did not temper the headlines.

The Sex Survey and U. S. Society Today

24Is there a way to think about the results from the survey itself (and the similar findings in France and England) that can help us either substantively or methodologically in trying to think about sex and gender as a broader cultural phenomenon?

25I have pondered this especially because I see such a wide chasm between this sort of recent research work in cultural studies, especially work on gender and gay and lesbian studies — work I expect will be featured at this conference — and social science research such as this survey. I would like to conclude by briefly addressing two different issues: 1) a few words about the framework that we tried to provide for thinking about these results and 2) comment on the role of empirical social science and some concerns I have about the reaction of both the "right" and the "left" to this research.

26Probably the most frustrating aspect of the process since the publication of the books is how little success we have had in communicating our understanding of the findings, and especially the social scientific framework that we worked so hard to formulate, especially in the academic book: The Social Organization of Sexuality (Laumann et al., 1994). While recognizing that there is a certain amount of distortion in the data from the interviews, mainly in the form of underreporting, the first step in any serious empirically based analysis is to take our results, that is the self-reports of a representative sample of the population, seriously. Self-report is not the same as behavior, but it is not wholly independent from it either.

  • 7 An example might help explain what I mean by judicious use of data. Many skeptics point to the fac (...)

27Distortion is also not purely random and answers to questions can be used judiciously as a proxy for direct observation and measures of behavior.7 What sense do we make of our somewhat surprising findings? First, they point to the disjunction between the ways that we talk about sex and the behavior itself.

28Our sociological approach emphasizes that sex is embedded in a complex set of structures: work, marriage, family, and leisure. It is embedded in ongoing relationships and networks and for most people most of the time, it is highly constrained and routinized. Acquisition of new partners, having many sexual encounters with a lot of variation are sustained only with great difficulty, for most people. One way to think about the contrast between all the sex talk and sex imagery in the media is as a kind of social fantasy life that stands in contrast with the more common everyday experience.

29On the other hand, there is an opposite kind of danger in overly simple generalizations about a large and varied population. While useful and important for certain purposes, summary statistics — averages and single percentages — that amalgamate the whole population mask the marked variation that can exist between groups. One very reasonable way to view the results from our study is as a description of quite varied and multiple social worlds. I am quite struck by how these worlds manage to coexist in modern Western societies. A very large section of American society — especially older, high school educated, religious, suburban and smaller town residents — live lives and hold views that are very traditional — one is tempted to say old-fashioned, almost from another world or era. This group accounts for one-third or more of the U. S. adult population. The obverse exists as well, younger persons (including large segments of persons in their 40s who came of age in the 1960s and 1970s), the more highly educated, urbanized population, especially when single either before settling down in a longer-term relationship or when between relationships, is much more experimental, much more active sexually, much more modern or even post-modern in its view. The extreme end of this distribution appears from the survey to represent a very small proportion of the whole population — maybe as little as a few percent. Of course, a single percent of the survey represents about 1. 5 million people. But I find that much of our thinking about changes in society, developments such as the sexual revolution or the gay and lesbian movement (not to mention the avant-garde of a movement such as contemporary "queer" culture) may be based on very small segments of the population. One of my favorite tactics in explaining this in the gay community, which in the United States is extremely wedded to the idea that 10 percent of the population is gay, is to point out to them that Jews make up about 2 percent of the U. S. population. Very few people in the U. S., Jewish or gentile, seem to know this. It just does not seem to be germane. Unlike the gay community, Jews are not thought of as (nor are they worried about) being powerless or lacking political or cultural influence.

30Political power and social influence is not merely an issue of numbers and percentages. A general survey such as ours tends to emphasize larger masses within the population and for that reason to emphasize the middle and the average. Contrasting that part of the population with the margins or the extremes both within the survey, but also as presented in certain media representations and in certain avant garde and left-wing discourses can help to explain some recent events (and some long-standing tensions) in the U. S. I'm thinking particularly of the political successes of the right wing, both in electoral victories in congress under the leadership of Newt Gingrich, and in grass-roots campaigns. Much of their rhetoric about "family values" and decrying moral decay is based on contrasting the everyday experience of large segments of the population (white, suburban, religious) with the very differently organized lives of the most visible "radical" elements.

31My final concern — speaking as a social scientist — but also a person concerned about social change, policy, and politics — and also as some one impressed with the efflorescence of cultural studies, particularly feminist and gay and lesbian studies — has to with tendencies I see on both the "right" and the "left" to reject quite completely and quite out of hand the results of studies such as ours. This has been most recently brought home to me by a major review of our research by a left-wing population geneticist from Harvard, Richard Lewontin, in the New York Review of Books, one of the most widely read general intellectual publications in the U. S. In "Sex, Lies, and Social Science," Lewontin argues that all self report data is fundamentally flawed and that therefore no useful conclusions can be drawn from our survey (Lewontin, 1995). His argument is so general as to lead to the rejection of all quantitative social science but I think the fact that the subject is sex makes his argument much more widely accepted. Strangely enough Lewontin himself seems to have no major substantive disagreement with our findings, especially not any disagreement about their magnitude. This is not someone who believes there is lots more sexual activity in American society than we found, if anything he argues that there is less. But drawing on important insights from post-structuralist thinking on narrative and social construction, he rejects the usefulness of this sort of research for any purpose.

32This is not very different from the reaction of the right-wing. As I was leaving the U. S., the social science e-mail networks were abuzz with a major impetus to cut the funding for the social sciences in the current round of budget cuts. The attitude appears to be that science and technology should be preserved, but social science can just be axed. Right wingers asked about the media coverage of the sex results when they first appeared, generally murmured approval, but always insisted that the government should not support such research.

33These sorts of attitudes, perhaps expected from the right, surprise and concern me when coming from the left. I think we (intellectuals generally, and progressive intellectuals in particular) miss an opportunity to try to rethink some of our pre-conceptions and assumptions about U. S. (and French and English) society when we reject statistical surveys. The point of empirical data collection in the sciences is to test intuitions and hypotheses. If we merely reject the data when it does not conform to our expectations, we avoid an important theoretical exercise. The great strength of surveys is that they are extremely "democratic" in their approach. Their greatest strength methodologically is in talking to a representative sample of the population. I am not saying that they give full or simple voice to everyone and to the full range of respondents thoughts, opinions, and feelings. The selection of topics and the framing of questions rests with the researchers — but the discipline of this type of research is that an unbiased sample of a clearly defined population is allowed, even invited to report on certain aspects of their experience. We ignore such reports at our peril if in fact we hope to better understand the worlds we live in.



Altman, Dennis, 1986. AIDS in the Mind of America. Garden City, New York: Anchor Press/Doubleday.

Crimp, Douglas, 1987. "How to Have Promiscuity in an Epidemic". October, 43, 237-271.

Davis, James A. and Tom W. Smith, 1988. General Social Surveys, 1972-1988: Cumulative Codebook. Chicago: National Opinion Research Center.

Elmer-Dewitt, Philip, 1994. "Now for the Truth About Americans and Sex: The first comprehensive survey since Kinsey smashes some our most intimate myths". Time, Vol. 144, no. 16. October 17, 62-70.

Institute of Medicine, 1986. Confronting AIDS: Directions for Public Health, Health Care, and Research. Washington, D. C.: National Academy Press.

Irvine, Janice, 1990. Disorders of Desire: Sex and Gender in Modern American Sexology. Philadelphia: Temple University Press.

Kinsey, Alfred C., Wardell B. Pomeroy, and Clyde E. Martin, 1948. Sexual Behavior in the Human Male. Philadelphia: Saunders.

Kinsey, Alfred C., Wardell B. Pomeroy, Clyde E. Martin, and Paul H. Gebhard, 1953. Sexual Behavior in the Human Female. Philadelphia: Saunders.

Laumann, Edward O., John H. Gagnon, Robert T. Michael, and Stuart Michaels, 1994. The Social Organization of Sexuality: Sexual Practices in the United States. Chicago: University of Chicago Press.

Lewontin, Richard, 1995. "Sex, Lies, and Social Science". The New York Review of Books. Vol. 42, no. 7, 20 Apr.

Michael, Robert T., John H. Gagnon, Edward O. Laumann, and Gina Kolata, 1994. Sex in America: A Definitive Survey. New York: Little Brown.

Patton, Cindy, 1985. Sex and Germs: The Politics of AIDS. Boston: The South End Press.

Shilts, Randy, 1987. And the Band Played On: Politics, People, and the AIDS Epidemic. New York: St. Martins Press.


1 This process is described in a series of publications from the National Academy of Sciences in Washington, a semi-official agency chartered by Congress. This initial work is documented in Confronting AIDS: Directions for Public Health, Health Care, and Research (Institute of Medicine, 1986).

2 This process with its peculiarly American emphasis on voluntarism and self organization, in the face of homophobic stigma and neglect by the government, building on and expanding existing institutions of a recently politicized gay community has been documented in a number of places. See especially Altman (1986), Shilts (1987), Crimp (1987), and Patton (1985).

3 The epidemiological models of disease prior to AIDS were not rooted in the notion of networks, but rather were based on air borne or insect borne diseases which are transmitted almost randomly through a population. AIDS and other STDs depend on intimate rather than casual contacts and these contacts are anything but random.

4 One of our key findings is that "homophily," the principle of "like attracts like" that shapes many social networks turns out to be widely applicable to sexual networks as well, including the formation of casual partnerships. As we learned to explain to the media looking for pithy sound bites, while we all may like the idea of falling in love across a crowded room, the persons in most crowded rooms are highly selected in terms of their social attributes.

5 We did not use the term heterosexual, instead we asked about the first time the respondent had vaginal intercourse. We did specify with a male or female, i. e., the opposite sex of the respondent. This is an example of the general strategy used throughout the questionnaire. Avoid scientific and loaded terminology to the degree possible. Ask questions as simply, directly, and specifically as possible.

6 In addition, about 20 percent of the women say they had intercourse with a man for the first time because it was their wedding night compared to only 7 percent of the men. Similar patterns exist for the men and women who report going along with their first heterosexual experience rather than "wanting" it. Except that among this group, about one quarter of both the men and women report "peer pressure" as being the major reason.

7 An example might help explain what I mean by judicious use of data. Many skeptics point to the fact that in our sample men report more female sexual partners than women report male sexual partners. In the whole population of persons who have sex with each other this cannot be the case. It turns out that the most plausible explanation for most of this difference is that there is some exaggeration in the men's reports and some minimization in the women's reports. Rather than leading us to conclude we cannot believe anything respondents say, we are led to conclude that our estimates are in the general ballpark. Interestingly enough men and women's reports of how often they have sex hardly differ at all, indicating that when there are less clear social norms, reports are less subject to distortion.


A sociologist, was the project director of the U. S.. National Health and Social Life Survey. As a co-author of The Social Organization of Sexuality : Sexual Practices in the United States, his primary concerns included methodological topics and homosexuality. He continues to do research at the University of Chicago on the sociological and cultural aspects of sexuality.

© Presses universitaires François-Rabelais, 1997

Conditions d’utilisation :

Cette publication numérique est issue d’un traitement automatique par reconnaissance optique de caractères.


Rechercher dans OpenEdition Search

Vous allez être redirigé vers OpenEdition Search