Lies, Damned Lies, and Statistics

It's unclear to me whether Jonathan is presenting statistics he knows to be faulty, thinking I could be easily misled, or if he has been duped by a flagrant abuse of numbers. In either case, here's what he has to say:

With regards to your "practical reasons" for keeping incest and polyganous marriage illegal, you should also consider adding homosexuality to that list if you wish to follow that type of reasoning.
1) homosexuals are not immune from abuse and rape
2) homosexual marriage does not produce birth defects, but the statistics show that homosexual relations is a dangerous health risk:

To point 1, unless he's prepared to argue that homosexuals are more abusive and more inclined to rapism than heterosexuals, this is utterly superfluous. To point 2, the statistics "show" only that he has no capacity for statistical analysis.

- "Of the 298,248 men (of 13 years or older) who were living with AIDS 57% were men who had sex with men (MSM)" - CDC (http://www.avert.org/statsum.htm).

Strictly speaking, this is untrue because the incidence of AIDS among men worldwide is many orders of magnitude higher among heterosexuals than homosexuals. That's because HIV does not discriminate. Sexual intercourse with an infected individual causes the spread of AIDS, not sex with gay men. Perhaps the HIV-positive populations of southern Africa have been secretly gay all along, but I find that unlikely. I suppose Jonathan meant to restrict his claim to the Unites States, where (here's a shocker) gay men have constituted a disproportionately high number of AIDS cases. The 57% figure, however, is quite interesting. What would that figure have been in 1986? 1990? Clearly, the rates of infection among gays relative to straights is decreasing, and this is not a surprise given the aggressive efforts at promoting safer sexual practices among gays over the last decade.

- "Homosexuals account for 3-4% of all gonorrhea cases, 60% of all syphilis cases, and 17% of all hospital admissions (other than for STDs) in the United States" despite only making up 1-2 percent of the total population. ("Changes in Sexual Behavior and Incidence of Gonorrhea." Lancet, April 25, 1987)

Whom does Jonathan expect to fool by citing statistics from early 1987? Even the percentage of the population that gays account for in this tidbit is wrong. As a percentage, it's anywhere from 5 to 10 times more than that. High rates of unprotected sex in any community will contribute to an increase in the spread of STD's. So what? The AIDS epidemic forced gays to practice safe sex. Their geometric rise in social visibility also enabled them to take better care of their health. Furthermore, I wouldn't cite statistics from the peak years of the bubonic plague as "proof" that being a European Christian is "a dangerous health risk."

-"The median age of death of homosexuals is 42 (only 9% live past age 65). This drops to 39 if the cause of death is AIDS. The median age of death of a married heterosexual man is 75." (numerous studies)

This "statistic" is an outrageous lie propagated by nutball anti-gay groups that has been debunked many, many times. (I note that, unlike his other citations, this one is only "numerous studies.") I wish I had the relevant articles on hand, because there are reams of them. Andrew Sullivan, in particular, has some really good stuff on this. Let me try to explain the shell game in fairly general terms. First of all, any sociological study of the gay population has to face the immediate problem that there are simply no data older than a couple, maybe three decades. What the people behind these studies did was to isolate the gay population that was still relatively young in the 1980s, include only homosexuals that had died in their analysis, and take the median age of those people. They failed or didn't want to take into account the fact that the vast majority of homosexuals in that age group were still alive and could be expected to live many more years.

Imagine taking a random sampling of one million Americans born during the 1960s. Say it's 1989, so their ages range from 20-29. Now exclude all individuals still alive in 1989, counting only those who had died. Now take their median age of death. The number you'd arrive at would be, for obvious reasons, outrageously low and also utterly meaningless, the result of a masturbatory exercise in number manipulation.

- The Medical Institute of Sexual Health reports [ Executive Summary, "Health Implications Associated with Homosexuality," 1999]: - "Homosexual men are at significantly increased risk of HIV/AIDS, hepatitis, anal cancer, gonorrhea and gastrointestinal infections as a result of their sexual practices." - "Women who have sex with women are at significantly increased risk of bacterial vaginosis, breast cancer and ovarian cancer than are heterosexual women." - "Domestic violence is...probably more common among homosexuals than among heterosexuals." - "Significantly higher percentages of homosexual men and women abuse drugs, alcohol and tobacco than do heterosexuals."

What, pray tell, is the Medical Institute of Sexual Health? No government agency publishes reports on "health implications associated with homosexuality," so I assume this is some sort of private foundation. I'd be willing to make a wager as to its credibility. The giveaway is the idea that homosexuality is a factor in cancer rates. This is obviously and flagrantly untrue to anyone who has a rudimentary knowledge of what cancer is, how it develops, and the mechanisms by which it spreads. Cancer is the result of a mutation of control genes in vulnerable parts of the body. It arises either out of random mutation or through direct damage to one's DNA. Sexual behavior has many implications for overall health, but it has absolutely nothing to do with cancer. Finally, the closing two claims are sociological and not medical, and rather difficult to comment on, since Jonathan includes neither a link nor any supporting data. The "probably more common" construction is particularly weaselly. Nevertheless, I would be willing to bet that these assertions are either ripped out of context or else pure fabrications. I report, you decide.