Point By Point: A Look At Paul Cameron’s Medical Consequences of What Homosexuals Do
Part 1: “What Homosexuals Do”
Laying the Groundwork
Jim Burroway "
January 31, 2006; revised October 15, 2006
Throughout Medical Consequences, Dr. Cameron describes various sexual practices in very explicit detail. But first, to avoid offending right away the sensibilities of an audience that’s almost certainly uncomfortable with homosexuality, he gently eases into this delicate subject by offering a broad overview of these sexual practices, which he summarizes in a table entitled “Homosexual Activities”:
|Fecal sex - eating||4||8|
|Sex with minors||37||23||24/|
After a few brief introductory paragraphs, Dr. Cameron directs the reader’s attention to that table while noting:
…two things stand out 1) homosexuals behave similarly world-over, and 2) as Harvard Medical Professor, Dr. William Haseltine,33 noted in 1993, the “changes in sexual behavior that have been reported to have occurred in some groups have proved, for the most part, to be transient”
13. Jay, Karla; Young, Allen. The Gay Report: Lesbians and Gay Men Speak Out About Sexual Experiences and Lifestyles (New York: Summit, 1977).
16. Gebhard, Paul; Johnson, Allen. The Kinsey Data: Marginal Tabulations of he 1938-1963 Interviews Conducted by the Institute for Sex Research (New York: Saunders, 1979).
18. Jaffe, Harold W.; Keewhan, Choi; Thomas, Pauline A.; Haverkos, Harry W.; Auerbach, David M.; et al. “National case-control study of Kaposi’s sarcoma and Pneumocystis carinii pneumonia in homosexual men: Part 1, Epidemiologic results.” Annals Of Internal Medicine 99, no. 2 (August 1983): 145-151.
19. Quinn, Thomas C.; Stamm, Walter E.; Goodell, Steven E.; Mkrtichian, Emanuel; Benedetti, Jacqueline; Corey, Lawrence; Schuffler, Michael D.; Holmes, King K. “The polymicrobial origin of intestinal infection in homosexual men.” New England Journal of Medicine 309, no. 10 (September 8, 1983): 576-582.
20. Biggar, Robert J.; Melbye, Mads; Ebbesen, Peter; Mann, Dean L.; Goedert, James J.; Weinstock, Robert; Strong, Douglas M.; Blattner, William A. “Low T-lymphocyte ratios in homosexual men: Epidemioligic evidence for a transmissible agent.” Journal of The American Medical Association 251, no. 11 (March 15, 1984): 1441-1446.
Rigdon, John E. “Overcoming a deep-rooted reluctance, more firms advertise to gay community.” The Wall Street Journal (July 18, 1991): B-1.
26. Elford, Jonathan; Tindall, Brett; Sharkey, Terry. Letter to the Editor: “Kaposi’s sarcoma and insertive rimming.” Lancet 339, no. 8798 (April 11, 1992): 938.
27. Beral, Valerie; Bull, Diana; Darby, Sarah; Weller, Ian; Carne, Chris; Beecham, Mary; Jaffe, Harold. “Risk of Kaposi’s sarcoma and sexual practices associated with faecal contact in homosexual or bisexual men with AIDS.” Lancet 339, no. 8793 (March 14, 1992): 632-635.
33. “AIDS Prognosis.” Washington Times (February 13, 1993) C1.
We’ve barely begun looking into Dr. Cameron’s pamphlet, and already we are confronted with several fundamentals problem with the “Homosexual Activities” table. First, it describes sexual activities for gay men only — he completely ignores lesbians. In fact, as we go through this pamphlet, we will see that lesbians are essentially invisible, which is a pretty glaring omission for a pamphlet that intends to describe “what homosexuals do”.
The second problem with the table is the way in which he arbitrarily groups unrelated activities together (in this case, toys with fisting, and public sex with orgy sex). It’s like mixing apples with oranges, and it makes unclear exactly which statistic is supposed to apply to which activity.
But that’s not all. The particular version of the table posted on the web suffers from many severe formatting errors, making it practically illegible. Fortunately, a similar table was printed in Dr. Cameron’s 1993 book The Gay Nineties.D Using that table as guidance, I’ve been able to determine that in the online version the data for “Fisting/Toys” is misaligned. I’ve shown the original misalignment, with the corrected alignment immediately below.
These formatting errors are very odd because the fundamental methods for creating web pages make formatting tables exceptionally easy — it doesn’t take a technical whiz to do this. Most training for creating web pages covers tables in the very early elementary stages, and just about all of the common web site creation tools make creating tables a snap. Tables are in integral part of how many web pages are formatted. If Dr. Cameron were really interested in the accuracy of his work, surely he could invest a little time or money to have this simple problem fixed. But already it provides us our first clue that carelessness will be one of the many hallmarks of this pamphlet.
Let’s examine each column of the table one by one:
Column 1: US, 1940s. Gebhard & Johnson (#16) was based on the pioneering surveys conducted by Alfred Kinsey and his associates beginning in 1938. This effort led to the publication of Sexual Behavior in the Human Male, which appeared in 1948 amid tremendous controversy. It was followed in 1953 by Sexual Behavior in the Human Female. After Kinsey’s death in 1956, the Kinsey Institute continued his surveys under the direction of Dr. Paul H. Gebhard until 1963, culminating in Gebhard’s and Alan Johnson’s The Kinsey Data: Marginal Tabulations of the 1938-1963 Interviews Conducted by the Institute for Sex Research in 1979.
The Kinsey surveys relied on volunteers rather than a statistically random sample. Because of the difficulty in finding homosexuals in the repressive atmosphere of the time, Dr. Alfred Kinsey relied on interviews of those found in a few large cities and members of homophile groups like the Mattachine Society. He also interviewed prisoners and other institutional populations. As a result, the homosexual sample ended up consisting of delinquent and non-delinquent subgroups, in the following numbers:E
Throughout the Kinsey reports, all of the the homosexual tabulations were divided between non-delinquent and delinquent categories, and the data was kept separate between the two groups. Kinsey did this because the delinquent and institutional populations were exceptionally unrepresentative in their sexual behavior.
This is important because when Dr. Cameron presents data from the Kinsey surveys, he combines the responses of the delinquent populations with the non-delinquent population. And he does this without revealing what he’s doing. This is deceptive because we know that many prisoners behave homosexually because that is the only outlet available to them. We also know that sex is often an expression of power and dominance in prison populations. But with Dr. Cameron simply combining the results from the two categories, he creates misleading statistics which are drastically skewed by the behaviors of an unrepresentative population.
This isn’t the only time Dr. Cameron has misused Kinsey’s data this way. In the mid-1980’s, he provided sworn statements based on a similar misuse of the Kinsey report to the U.S. District Court in Dallas for Baker vs Wade (an attempt to overturn Texas’ sodomy law). But Judge Jerry Bruchmeyer, in his ruling on the case, found that
Dr. Paul Cameron... has himself made misrepresentations to this Court. For example:
(i) his sworn statement that "homosexuals are approximately 43 times more apt to commit crimes than is the general population" is a total distortion of the Kinsey data upon which he relies — which, as is obvious to anyone who reads the report, concerns data from a non-representative sample of delinquent homosexuals (and Dr. Cameron compares this group to college and non-college heterosexuals);
(ii) his sworn statement that "homosexuals abuse children at a proportionately greater incident than do heterosexuals" is based upon the same distorted data — and, the Court notes, is directly contrary to other evidence presented at trial.F
Gebhard & Johnson themselves cite several other problems with the data in the Kinsey surveys.G They noted that Kinsey didn’t trust probability sampling methods, which were in their infancy in those days. He also didn’t keep records of refusals, which means there is no way of knowing the response rate. And in his later years, as criticisms mounted over his methods and conclusions, he became overly sensitive to suggestions which might have improved the quality of his findings (for example, omitting prison populations from his tabulations).
In the end, the Kinsey surveys are a landmark effort in what was an emerging study of human sexuality. It blazed a trail through uncharted territory, pioneering a whole knew way of studying human behavior. Like all such groundbreaking efforts, it is primitive by today’s standards in many important ways. While it lacked the statistical rigor that is expected of today’s surveys, it was an important laboratory for developing many of the interviewing techniques used in later surveys. It also provided a tantalizing glimpse into a subject that had never before been exposed to academic examination — the intensely private and intimate lives of ordinary people. In this way, the Kinsey surveys opened a window to a whole new field in the social sciences. But its value is found more in the establishment of a new field in the social sciences, not in the numerical statistics of the survey itself.
Column 2: US, 1977. Jay & Young (#13) is modeled after many other informal sex surveys which were popular in the 1970’s and 1980’s. Magazines like Cosmopolitan, Red book and Playboy often conducted reader surveys, and books like The Hite Report on Male Sexuality were best-sellers, satisfying a voracious demand among a curious audience during the anything-goes decade of the 1970’s. The Gay Report, a book by Karla Jay and Allen Young, was one such informal survey. These surveys were criticized for being woefully inaccurate, which is why they are rarely performed anymore. The problems with Jay & Young’s book typify these shortcomings.
The Gay Report
One of the most popular sources of statistics for anti-gay activists is Karla Jay and Allan Young’s The Gay Report. This survey was the culmination of 5,400 responses to an exceptionally large casual survey. But in reading the report, it quickly becomes apparent that size doesn’t matter much when it comes to accuracy. For more information, see our review of The Gay Report.
To prepare for this study, the authors mailed “hundreds of thousands” of 16-page surveys across the country, with an abridged version printed in the adult magazine Blueboy, which is sort of a gay equivalent to Playboy or Penthouse. But this massive distribution elicited only 5,400 responses. The authors estimate that some 500,000 gays and lesbians saw the questionnaire, which makes the response rate would be an abysmally low 1%.
Surveys with such an incredibly low response rate are essentially worthless from a statistical standpoint. Critics are quick to question what this survey can say about the more than 98% who, when confronted with a 16-page questionnaire asking personal and explicit questions about their sexual experiences, declined to answer. And when some respondents commented that it took from six hours to three days to complete it,H it raises an even more important question: what does it say about the very few who did respond?
Most of those who respond to casual surveys like this one do so because they feel they have something unusual or interesting to share, a phenomenon known as “participation bias.”I Such participants tend to be more sexually experienced, more interested in sexual variety, and less inclined to follow rules or conform to social norms.J They are generally more eager to share their unusual or “kinky” sexual behavior with a larger audience. Conversely, those who choose not to respond tend to be more conservative and less sexually experienced.K
What’s more, research has shown that if there is an implied approval of some of the activities asked about in sexuality surveys, those who are more adventurous are much more likely to share their experiences.L And there is considerable evidence throughout this book that such biases exist — not only in how the questions were worded, but in how the authors interpreted of the results.
And there’s one other point to keep in mind. Nearly 45% of all responses came from Blueboy readers. Surveys of magazine readers, at best, reflect only the attitudes of those who read that magazine. And since Blueboy is a gay “adult” magazine, it should come as no surprise that the responses from 2,462 Blueboy readers would skew the results dramatically. With Blueboy readers making up almost half of all respondents, this study would be no more representative of the overall gay population than would a study half made up of Playboy readers represent the overall straight population
While the authors expressed great pride in their work, they also recognized at least a few of their limitations. On page 10 they grudgingly say that they “do not claim to have a scientific or representative sample of lesbians and gay men.”
Column 3: US, 1983/84. The data in this column is unattributed. The poor formatting online at one point led me to believe that it was referring to a letter to the editor that he references later in the pamphlet. But the clue, once again, came from his book, The Gay Nineties, where Dr. Cameron reprinted much of an earlier version of Medical Consequences, including an earlier version of this table. At one point in the book, he referred to some of the statistics corresponding to column three while referring to “the largest random survey of gays.”
This phrase is the one he often uses to describe his own deeply flawed sexuality survey (sometimes known as the ISIS Survey) that he conducted in 1983 and 1984. I looked at that survey, and sure enough the data matched column three of this table.M
You can read ahead and learn more about Paul Cameron’s ISIS Survey here.
There are many problems with this survey, too many to go into here. We will explore this survey in much greater depth in the next section. But for now, the most important thing to keep in mind is that it shared at least one problem with The Gay Report: its exceptionally low response rate (23.6%). It also yielded an exceptionally low number of gay men (41, with 39 bisexuals) and lesbians (25, with 44 bisexuals), which makes the claim of being “the largest random survey of gays” rather suspect, to say the least.
Column 4: US, 1983. Jaffe, et al. (#18), a very early AIDS study, is fascinating to read in light of what we now know about the disease. The lead author, Harold Jaffe, was a member of the Centers for Disease Control and Prevention’s (CDC) Kaposi’ Sarcoma and Opportunistic Infections Task Force, which was formed in early 1981 to investigate the mysterious deaths of five gay men in Los Angeles over the previous six months. Dr. Jaffe worked heroically, despite the appalling lack of funding in the first full year of the epidemic, to investigate what everyone would later come to know as AIDS. In that capacity, he was just one of a very few witnesses to history, those dedicated few who investigated the growing number of reports to the CDC of gay men suffering from an array of mystifying symptoms.
Dr. Jaffe was among the first to suspect that this new disease may be sexually transmitted. Oddly, this notion was dismissed by most other researchers, who believed that the immune systems of these gay men were being compromised by drug use or the presence of semen. These beliefs held sway, despite the fact that heterosexual women weren’t experiencing immune deficiencies, and drug use had been common in society for decades with no similar problems. This study was among the very early ones to demonstrate that AIDS really was a sexually transmitted disease (STD), although they weren’t able to correctly identify how AIDS was transmitted. That would come later.
In order to make the connection between AIDS and sexual activity, the researchers recruited two groups of participants. The first group (50 men) consisted of patients with Pneumocystis pneumonia (a rare form of pneumonia common among early AIDS patients), Kaposi’s sarcoma (a rare form of skin cancer also common among early AIDS patients), or both. These diseases were selected because they were the most common indications for doctors to rely on to know whether a patient had AIDS, especially before reliable HIV tests became available.
The second group, the control group, consisted of 120 men who had no symptoms, and thus were presumed to be disease-free. But two-thirds of this control group were recruited from STD clinics as well. The researchers did this because “obtaining a true random sample of homosexual men to serve as controls did not appear feasible.” In the end, more than three-quarters of the combined study population was either sick with AIDS or recruited from an STD clinic, and much of the control group ended up showing symptoms of AIDS before the study was finished.
When researchers target high-risk groups of men for their studies, it should come as no surprise that their reports describe men engaging in risky behavior. Data from these studies ignores healthy gay men. No study in which participants are drawn mostly from STD clinics would ever come close to representing the general gay male population. But Dr. Cameron has no qualms about using this to make claims about the sexual behavior of all gay men.
Column 5: Denmark, 1984. Biggar, et al. (#20) is another pioneering AIDS study, one that the authors had a great deal of difficulty convincing a journal to publish. Its conclusion was that AIDS was caused by an infectious agent, at a time when the medical community was still biased towards the unproven belief that AIDS was caused by drug use or the presence of semen.
This study was performed with a convenience sample recruited from STD clinics. The authors clearly state that the study consists of 132 gay men in the cities of Aarhus and Copenhagen. Only 80 returned for the follow-up study and two were dismissed due to various problems, resulting in a 59% response rate. The authors also make clear that “the participants were specifically selected as to their life-style and were unlikely to be representative of the general homosexual community of either city.” With statements like this in plain English, it is unconscionable to use this data to characterize the behaviors of the overall gay male population.
In the same footnote, Dr, Cameron also cites the July 18, 1991 Wall Street Journal. The article, which reports on mainstream businesses advertising in gay magazines, isn’t relevant to this table. It’s doubtful that Dr. Cameron intended to use this Wall Street Journal article to support this data. Later in the pamphlet, he will cite this footnote again where the Wall Street Journal article is relevant. It would have been clearer if he had presented this source as a separate footnote and cited it only when it was relevant.
Column 6: US, 1983. Quinn, et al. (#19) is yet another early AIDS study with a convenience sample of 194 homosexual or bisexual men, each and every one a patient at an STD clinic. One hundred and nineteen of these patients were selected because they “had an acute illness of less than four weeks’ duration and had not taken antibiotics during the two weeks before enrollment” — in other words, they were selected because they were sick. This study didn’t try to characterize sexual practices of gay or bisexual men in the general population.
Column 7: London, 1985. Beral, et al. (#27), is still another early AIDS study. This one began in 1982, even though this article wasn’t published until ten years later. This convenience sample consisted of only 65 participants who were not only selected from an STD clinic, but they were further chosen according to their particularly high-risk behavior which led to their being diagnosed with AIDS. The authors note that these men “were in the forefront of the AIDS epidemic in the U.K. and tended to be highly sexually active.” Since this study population was screened according to their high risk behavior, it cannot be construed as representing of the gay male population.
Column 8: Sydney/London, 1991. Elford, et al. (#26) is improperly cited in the pamphlet. Dr. Cameron should have noted that it is a letter to the editor (I have corrected the citation here). While this omission isn’t necessarily unusual in popular publications, it goes against the guidance of the major style guides for professional research papers and journals. This is important because letters to the editor are not peer reviewed. By omitting this fact in his citation, the reader may be left with the impression that it is a peer-reviewed article, which would be misleading.
The letter, only four paragraphs long, doesn’t describe how they selected the study study participants, nor does it indicate that the study population was intended to be representative of the general gay population.
So, what do we have? Of the eight sets of data in the “Homosexual Activities” table, only two were general population surveys: Kinsey’s survey (where Dr. Cameron improperly combined the delinquent and non-delinquent data), and Jay & Young’s book (with its abysmally low response rate of less than 5%). Fully half of the data in this table comes from early AIDS studies consisting of patients recruited from STD clinics. Two aren’t even peer-reviewed studies at all, but are letters to the editor. Consequently, this table is utterly worthless in characterizing the sexual behaviors of the overall gay male population.
Finally, much has changed since Dr. Hazeltine’s 1993 statement that changes in behavior are temporary. The Centers for Disease Control and Prevention (CDC) report more than a 45% drop in annual AIDS diagnoses since the peak in 1993.N Chicago reports a drop of nearly half since the mid-nineties.O New YorkP and Los AngelesQ experienced more than a 60% drop. In San Francisco, the drop has been even more dramatic — they’ve experienced more than an 80% reduction of new AIDS cases since 1993.R Something clearly has changed. Relying on obsolete data, as Dr. Cameron does, fixes his arguments to the very peak of the AIDS crisis.
Perhaps the most striking thing in what we’ve seen so far is Dr. Cameron’s heavy dependence on AIDS studies. Responsible professionals recognize very well the dangers of trying to use studies drawn from STD clinics to characterize the behaviors of gay men in the general population. As one researcher pointed out:
Homosexual men who attend ATD clinics have higher risk behaviors than those who do not, and STD clinic surveys will therefore tend to overestimate the prevalence of these behaviors.”S
Paul Cameron vs. Professional Ethics
By now, you have a pretty good idea of what Paul Cameron’s “research” is all about. You can learn more about what other social scientists think about him in “Paul Cameron vs. Professional Ethics.”
This is a well-known problem that anybody even remotely familiar with the social sciences can understand. Dr. Cameron’s deliberate misuse of this data without explaining where it comes from is not only highly unethical, it represents a slander of the worst order. This misrepresentation has earned him more than a rebuke from a judge in Dallas. This same type of misconduct led to his being dropped from the American Psychological Association, and a rebuke from the Nebraska Psychological Association and the American Sociological Association.
Please continue with:
Part 2: “Oral Sex”, in which Dr. Cameron’s famous survey makes its appearance.
I. Fenton, Kevin A.; Johnson, Anne M.; McManus, Sally; Erens, Bob. “Measuring sexual behaviour: Methodological challenges in survey research.” Sexually Transmitted Infections 77, no. 2 (April 2001): 84-92. Full text is available online for free at http://sti.bmjjournals.com/cgi/content/full/77/2/84.
Strassberg, Donald S.; Lowe, Kristi. “Volunteer bias in sexuality research.” Archives of Sexual Behavior 24, no. 4 (August, 1995): 369-382.
L. Wiederman, Michael W.; Weis, David L.; Allgeier, Elizabeth Rice. “The effect of question preface on response rates to a telephone survey of sexual experience.” Archives of Sexual Behavior 23, no. 2 (April 1994): 203-215.
Cameron, Paul; Cameron, Kirk; Proctor, Kay. “Effect of homosexuality upon public health and social order.” Psychological Reports 64, no. 3 (June 1989): 1167-1179.
N. Centers for Disease Control and Prevention. HIV/AIDS Surveillance Report (2002) no. 14. (Atlanta: US Department of Health and Human Services, Centers for Disease Control and Prevention; 2003): Front Cover. Available online at http://www.cdc.gov/hiv/stats/hasr1402/2002SurveillanceReport.pdf.
Centers for Disease Control and Prevention. HIV/AIDS Surveillance Report (2004) no. 16. (Atlanta: US Department of Health and Human Services, Centers for Disease Control and Prevention; 2005): 12, 16. Available online at http://www.cdc.gov/hiv/stats/2004SurveillanceReport.pdf.
O. Chicago Department of Public Health. HIV/AIDS/STD Chicago (Chicago: HIV/AIDS Surveillance Program, Summer, 2005): 4. Available online at http://egov.cityofchicago.org/webportal/COCWebPortal/COC_EDITORIAL/AIDSChicagoSummer05.pdf.
P. New York City Department of Health and Mental Hygiene. NYC Health web site: HIV/AIDS Surveillance Statistics, 2004.; http://www.nyc.gov/html/doh/downloads/pdf/ah/surveillance2004_tables_all.pdf.
Q. Los Angeles Department of Health Services. HIV/AIDS Semi-Annual Surveillance Summary (Los Angeles: Los Angeles Department of Health Services; July 2005): 2. Available online at http://lapublichealth.org/wwwfiles/ph/hae/hiv/Semiannual_Surveillance_Summary_July_2005.pdf.
R. San Francisco Department of Public Health. HIV/AIDS Epidemiology Annual Report, 2004. (San Francisco: San Francisco Department of Public Health, 2003): 62. Available online at http://www.sfdph.org/PHP/RptsHIVAIDS/HIVAIDAnnlRpt2004-20050609-fnlWeb.pdf.
S. Fenton, Kevin A.; Johnson, Anne M.; McManus, Sally; Erens, Bob. “Measuring sexual behaviour: Methodological challenges in survey research.” Sexually Transmitted Infections 77, no. 2 (April 2001): 84-92. Full text is available online for free at http://sti.bmjjournals.com/cgi/content/full/77/2/84.